Method for Evaluating Crashworthiness of Railway Vehicles Based on Correlation with Injury Severity of Passengers Occupying Longitudinal Seats

Abstract

Improving safety for train passengers in the event of a collision is a priority. To this end, this paper proposes safety indexes. The severity of injury for these indexes was estimated for a level crossing accident using numerical simulation. We compared the correlation between the severity of injury of a human model and the safety index of vehicles, which is the integral of deceleration waveforms, mean deceleration waveforms and maximum deceleration waveforms. It was found that the integral of deceleration values had the highest correlation with the injury values. We therefore propose using the integral of deceleration as a method for evaluating crashworthiness design.

1. Introduction

In Japan, design standards for railcar body structures do not take into account collision accidents, and do not specify indexes for evaluating collision safety. Design standards for the collision safety of railcar bodies are specified however, in European countries and the United States. For example, in European countries, the average deceleration that occurs inside the car during a collision is used as an index, assuming several accident scenarios including a train-to-train collision (collision speed of 36 km/h) [1]. In the United States, the maximum deceleration that occurs inside the car during a train-to-train collision (collision speed of 48 km/h) is used as an index [2]. Each of these indexes has its own limit value. The assumed accident scenarios and the evaluation indexes vary depending on the railway systems and past accident cases in each country.

On closer investigation, it was found that out of all train accidents in Japan, train-to-train collisions are rare [3] compared to train-automobile collisions at level crossings that are the most common type of accident. Therefore, in Japan, collision with automobile should be a priority. Although there have been studies on the crashworthiness of car body structures in Japan [4, 5, 6, 7], these have focused on examining the vehicle's impact absorbing structure and reproducing the phenomenon through the analysis of collision tests. Consequently, there is limited knowledge that can be used to establish design standards.

In this report, we performed an injury analysis of passengers seated on longitudinal seats, in different conditions of varying train collision speed and collision position of trains. This analysis was based on the result of a statistical survey of major level crossing accidents that have occurred in Japan in the past. We also estimated the severity of head injuries in secondary collisions with a bench-end partition installed at the end of longitudinal seats (it has been confirmed that the risk of head injury is high in a secondary collision with bench-end partitions [3]). In addition, we present crashworthiness evaluation indexes and limit values for car body structures appropriate to the situation in Japan, by comparing the correlation between the “integral of deceleration” suggested by Okino et al. [8] for passengers seated on rotating reclining seats, and the Western “mean deceleration” and “maximum deceleration” indexes, which estimate injury levels. There have been no reports that have evaluated the safety of passenger compartments for passengers seated on longitudinal seats using Western indexes under various conditions that assume actual level crossing accidents and that have verified the validity of the evaluation indexes for car body structures using the degree of injury to passengers as a criterion. Therefore, verification results from this study will contribute to the establishment of design standards for the crashworthiness of railway vehicles in Japan.

2. Assessment of passenger compartment safety under various level crossing accident conditions

In this report, we carried out a safety evaluation by combining “level crossing accident analysis” and “passenger injury analysis” using numerical analysis. In the former numerical analysis, a train mathematical model was collided with a large dump truck mathematical model under various conditions, and the impact deceleration waveform generated in the train model was calculated. In the latter numerical analysis, a passenger model was seated in a longitudinal seat model, and the deceleration waveform obtained from the level crossing accident analysis was input to an interior equipment model including the longitudinal seat model to reproduce the interior situation of a train at the time of an accident.

For the “level crossing accident analysis” we used the model constructed by Okino et al. [9]. The analysis model consists of a train model with a standard stainless steel body structure and a large dump truck model (Fig. 1). Using this analysis model, the impact deceleration waveform (relative to the train travel direction) of the train passenger compartment when colliding with a dump truck was calculated using the input parameters of the train collision speed, left and right collision positions, collision angle, top and bottom collision positions and cargo mass. From these waveforms, the “mean deceleration” in accordance with European standards [1], the “maximum deceleration” in accordance with US standards [2], and the “integral of deceleration” proposed by Okino et al. [8] were calculated as evaluation indexes for the vehicle body structure.

Fig. 1　Train model and dump truck model

For the “passenger injury analysis,” we used a passenger injury analysis model [10] that we developed to estimate the injury level of passengers seated in longitudinal seats. This analysis model is composed of a human dummy model that imitates a passenger (a rigid model of the ES-2 dummy developed in the automotive industry to evaluate side impacts on the body) and a model of the longitudinal seats and bench-end partition installed at the ends of seats that are actually used in trains (Fig. 2). This model can output the level of head injury at the time of impact by inputting the cabin's impact deceleration waveform obtained from the above-mentioned “level crossing accident analysis.” Figure 2 shows an example of a dummy model sitting in the third seat from the bench-end partition, colliding with the partition and falling down.

Fig. 2　Dummy model and longitudinal seat model

By comparing the four evaluation indexes obtained from the “level crossing accident analysis” (two types of mean deceleration, maximum deceleration, and integral of deceleration) with the head injury severity obtained from the “passenger injury analysis,” we investigated evaluation indexes that were highly correlated with passenger injury severity.

2.1 Conditions and evaluation indexes for railroad crossing accident analysis

A total of 37 level crossing accident analysis scenarios were performed by combining the following factors: vehicle collision speed, collision position in the horizontal direction, collision angle, collision position in the vertical direction, and cargo mass (Table 1). As shown in Fig. 1, under the basic condition of Case 1-1, the horizontal collision position was set so that the centerline of the carbody was aligned with the center of the dump truck's load. The collision angle was set so that the vehicle and the dump truck were positioned at right angles to each other in the direction of travel. The vertical collision position was set so that the bottom of the underframe was 355 mm lower than the bed bottom of the dump truck (the reference height). The cargo mass was 11,000 kg. Cases 1-2 to 1-5 represent condition groups with different horizontal collision positions. Cases 2-1 to 2-4 represent condition groups with different collision angles, Cases 3-1 to 3-3 represent condition groups with different vertical collision positions, and Cases 4-1 to 4-5 represent condition groups with different cargo masses.

Table 1　Analysis conditions of level crossing accident

Analysis condition	Collision speed [km/h]	Collision position in horizontal direction	Collision angle [degree]	Collision position in vertical direction	Cargo mass [kg]
Case1-1_20 km/h	20	Center of load	0	Reference position	11,000
Case1-1_30 km/h	30	Center of load	0	Reference position	11,000
Case1-1_40 km/h	40	Center of load	0	Reference position	11,000
Case1-1_54 km/h	54	Center of load	0	Reference position	11,000
Case1-1_60 km/h	60	Center of load	0	Reference position	11,000
Case1-2_40 km/h	40	Center of gravity of dump truck	0	Reference position	11,000
Case1-3_40 km/h	40	Center of gravity of cabin	0	Reference position	11,000
Case1-4_40 km/h	40	1/2 lap	0	Reference position	11,000
Case1-5_40 km/h	40	1/4 lap	0	Reference position	11,000
Case2-1_40 km/h	40	Center of load	+5	Reference position	11,000
Case2-2_40 km/h	40	Center of load	−5	Reference position	11,000
Case2-3_40 km/h	40	Center of load	+10	Reference position	11,000
Case2-4_40 km/h	40	Center of load	−10	Reference position	11,000
Case3-1_40 km/h	40	Center of load	0	Reference position -177 mm	11,000
Case3-2_40 km/h	40	Center of load	0	Reference position -354 mm	11,000
Case3-3_40 km/h	40	Center of load	0	Reference position -512 mm	11,000
Case4-1_40 km/h	40	Center of load	0	Reference position	0
Case4-2_40 km/h	40	Center of load	0	Reference position	2,750
Case4-3_40 km/h	40	Center of load	0	Reference position	5,500
Case4-4_40 km/h	40	Center of load	0	Reference position	8,250
Case4-5_40 km/h	40	Center of load	0	Reference position	13,750
Case1-2_54 km/h	54	Center of gravity of dump truck	0	Reference position	11,000
Case1-3_54 km/h	54	Center of gravity of cabin	0	Reference position	11,000
Case1-4_54 km/h	54	1/2 lap	0	Reference position	11,000
Case1-5_54 km/h	54	1/4 lap	0	Reference position	11,000
Case2-1_54 km/h	54	Center of load	+5	Reference position	11,000
Case2-2_54 km/h	54	Center of load	−5	Reference position	11,000
Case2-3_54 km/h	54	Center of load	+10	Reference position	11,000
Case2-4_54 km/h	54	Center of load	−10	Reference position	11,000
Case3-1_54 km/h	54	Center of load	0	Reference position -177 mm	11,000
Case3-2_54 km/h	54	Center of load	0	Reference position -354 mm	11,000
Case3-3_54 km/h	54	Center of load	0	Reference position -512 mm	11,000
Case4-1_54 km/h	54	Center of load	0	Reference position	0
Case4-2_54 km/h	54	Center of load	0	Reference position	2,750
Case4-3_54 km/h	54	Center of load	0	Reference position	5,500
Case4-4_54 km/h	54	Center of load	0	Reference position	8,250
Case4-5_54 km/h	54	Center of load	0	Reference position	13,750

The collision speeds were set to five conditions: 20 km/h, 30 km/h, 40 km/h, 54 km/h, and 60 km/h. 54 km/h corresponds to the average estimated collision speed of serious accidents at level crossings occurred in Japan between FY1987 and FY2010 [11]. The horizontal collision positions were set to the following five conditions: the center of the load, the center of gravity of the entire dump truck including the load, the center of gravity of the dump truck cabin, the rear end of the load of the dump truck (overlap condition where half the width of the train body overlaps), and a position 750 mm rearward from the rear end of the load (overlap condition where 1/4 the width of the train body overlaps). The collision angle was defined as positive when the dump truck cab rotated towards the train, with the center of the load acting as the axis of rotation. The condition where the train and dump truck were perpendicular to each other was set to 0 degrees. And five conditions were set to ±5 degrees and ±10 degrees including this. Four vertical collision positions were set: the reference height, a position where the dump truck was 177 mm lower than the reference height, a position where the dump truck was 354 mm lower than the reference height (condition where the center height of the train's floor structure and the main frame of the dump truck's loading platform are aligned), and a position where the dump truck was 512 mm lower than the reference height (condition where the height of the underside of the train's floor structure and the underside of the dump truck's load are aligned). Six cargo mass conditions were set: 0 kg, 2,750 kg, 5,500 kg, 8,250 kg, 11,000 kg, and 13,750 kg (assuming a 25% overload).

To incorporate the effects of passenger position within the train, three locations within the cabin were evaluated for each scenario: the front, center, and rear. A total of 111 deceleration waveforms were output and four evaluation indexes were calculated from each waveform: two types of “mean deceleration,” “maximum deceleration,” and “integral of deceleration.” The two types of “mean deceleration” are indexes [1] that are the maximum value of the moving average of the deceleration waveform over 30 ms and 120 ms sections, hereafter referred to as “mean deceleration_30 ms” and “mean deceleration_120 ms” respectively. Additionally, “maximum deceleration” is an index [2] that is the maximum value of the deceleration waveform processed through a 50 Hz low-pass filter. The “integral of deceleration value” is the single integral of the deceleration waveform up to the time t_imp, at which point the double integral of the deceleration waveform becomes equal to the distance between the initial seating position of the dummy and the bench-end partition (see Fig. 3, seat 1: 277 mm, seat 2: 690 mm, seat 3: 1,150 mm, seat 4: 1,610 mm). In other words, t_imp corresponds to the estimated time until the passenger collides with the partition for a second time (secondary collision).

Fig. 3　Condition of initial seating position

2.2 Conditions and injury index for passenger injury analysis

To evaluate the effect of a passenger's initial seating position on injury levels during a secondary collision with the bench-end partition, we created a scenario in which one dummy model was seated in seats 1 to 4 relative to the partition (Fig. 3). The seat width for one person was set to 460 mm, and the center of the seat was positioned at the center of the head. However, for the first seat condition, the dummy model was too large to be seated in the center of the seat width, so it was seated 47 mm away from the center and the partition. By inputting the 111 deceleration waveforms described in the previous section for each of these seating conditions, a total of 444 conditions were analyzed.

The head injury level and head speed in the direction of train travel at the time of the secondary collision with the partition (secondary impact velocity) were calculated using the dummy model. The head injury level was calculated using the Head Performance Criterion (hereinafter referred to as HIC, as it is calculated in the same way as the Head Injury Criterion, which is an index of head injury from the front direction) specified in the Safety Standards [12]. The higher the value, the higher the risk of injury. The limit value is set at 1000. HIC was calculated from the resultant acceleration of the dummy head in three translational directions using eq. (1). In this report, injury severity was evaluated using HIC and its limit value of 1000.

  

$$\mathit{HIC}={\left\{\left(t_2-t_1\right){\left[\frac{{\int }_{t_1}^{t_2}a\left(t\right)\mathit{dt}}{\left(t_2{-t}_1\right)}\right]}^{2.5}\right\}}_{\mathit{max}}$$(1)

|t₁−t₂|≤36ms, a(t): Head translational

3-axis resultant acceleration [G]

3. Results of passenger injury analysis and evaluation indexes

Of the total 444 conditions, 375 conditions (1st seat: 81 conditions, 2nd seat: 111 conditions, 3rd seat: 111 conditions, 4th seat: 72 conditions) were confirmed to cause the head to undergo a secondary collision with the bench-end partition. Of these, 75 conditions were confirmed to cause the HIC to fall below 50 (5% of the limit value of 1000). In this report, we evaluated 300 conditions (1st seat: 62 conditions, 2nd seat: 111 conditions, 3rd seat: 87 conditions, 4th seat: 40 conditions) in which secondary impact behavior was confirmed and the HIC was 50 or higher.

3.1 Evaluation based on deceleration integral value

Figure 4 shows the relationship between the “deceleration integral value” and HIC for each initial seating position, as shown in Section 2.1, along with the limit value. In the first seating condition, the limit value was significantly lower than the limit value in all seating conditions. In the fourth seat, the limit value was exceeded in only one condition, while in the second seat, the limit value was exceeded in 77 conditions and in the third seat, the limit value was exceeded in 31 conditions. Within the range of accident scenarios considered in this study, the risk of head injury was higher in the second and third seats than in the first and fourth seats, with the second seat tending to be the highest. In general, it was confirmed that the HIC tended to increase as the deceleration integral value increased in all seating conditions, but in the second seat, it was confirmed that once the deceleration integral value exceeded approximately 5 m/s, the HIC did not increase any further and tended to decrease. It is thought that this is due to the dummy's head straddling the upper part of the bench-end partition immediately after the secondary impact. Because the behavior of the dummy affecting the HIC differs when the dummy is and isn't straddling (described in detail in Section 4.1), section 3.2 examined the evaluation indexes for the car body structure in the non-straddling condition.

Fig. 4　Comparison of HIC and integral of deceleration by the seating position

3.2 Study of evaluation indexes for vehicle body structure

In the second and third seat conditions, where the risk of head injury was high, the coefficients of determination were calculated using a linear approximation of four vehicle structure indexes, namely “integral of deceleration,” “mean deceleration at 30 ms” and “mean deceleration at 120 ms” in the European standard, and “maximum deceleration” in the US standard, and the HIC (an indicator of injury severity). And the correlation between these indexes and the HIC was compared (Fig. 5). For these comparisons, 78 conditions were used for the second seat, excluding the conditions in which the straddling behavior described in the previous section was observed from the 111 conditions in which the HIC was 50 or more, and 87 conditions in which the HIC was 50 or more were used for the third seat.

Fig. 5　Comparison of correlation between HIC and safety evaluation indexes at Ps2 and Ps3

Comparing the coefficients of determination for the four indexes and HIC (Fig. 5(e)), the coefficients of determination for “integral of deceleration” and “mean deceleration at 120 ms” were similarly high in the second seat condition, and the coefficient of determination for “integral of deceleration” was the highest in the third seat condition. This indicates that under the seating conditions in question, the “integral of deceleration value” is an index that has a higher correlation with the degree of injury to passengers than the “mean deceleration” and “maximum deceleration.”

4. Discussion

4.1 Considerations regarding seating conditions for the second seat

In the second seating condition, where the risk of head injury was the highest, the HIC tended to decrease when the “deceleration integral” exceeded approximately 5 m/s (Fig. 4). Previous sled tests (tests in which the “passenger injury analysis” performed in this report was performed using actual in-car equipment and dummies) confirmed that the HIC increases as the secondary impact velocity of the head increases [13]. Therefore, the secondary impact velocity and “deceleration integral” in the second seating condition were compared (Fig. 6). As shown in this figure, there was no tendency for the secondary impact velocity to decrease when the deceleration integral was 5 m/s or more, so it can be concluded that the decrease in HIC is not caused by a decrease in the secondary impact velocity.

Fig. 6　Comparison of SIVH and Integral of deceleration at Ps2

Comparing the “deceleration integral” and the dummy behavior as shown in Fig. 7, when the “deceleration integral” increased, the dummy's falling toward the bench-end partition side was reduced, and the head was confirmed to straddle the upper part of the partition immediately after the secondary impact. Specifically, as shown on the right side of Fig. 7, when the head hits the partition at a height of approximately 1.2 m or more from the floor, the head moves over the partition to the right side of the figure immediately after the secondary impact. This behavior reduces the impact acceleration generated at the head during the secondary impact, resulting in a lower HIC. This phenomenon has also been confirmed in sled tests [3].

Fig. 7　Head position from the floor at secondary impact vs. integral of deceleration

4.2 A study on the correlation between vehicle body structure evaluation indexes and injury severity

In the previous chapter, we clarified that the “integral of deceleration” was highly correlated with the degree of injury to passengers in the second and third seat conditions, which have a high risk of head injury. To investigate this further, we calculated the coefficient of determination by linearly approximating the “integral of deceleration,” “mean deceleration,” and “maximum deceleration,” which are evaluation indexes of the car body structure, as well as the secondary impact velocity. We then compared the correlation (Fig. 8). For the second seat condition, data from 78 conditions in which the head did not straddle the bench-end partition, as described in the previous section, was used. Figure 8(a) shows that the secondary impact velocity increases as the “integral of deceleration” increases. As confirmed in Section 3.1, the risk of injury is highest in the second seat and the secondary impact velocity was also high. Furthermore, the coefficient of determination for the “integral of deceleration” was high for both seating conditions (Fig. 8(e)), indicating that the secondary impact velocity, which affects the HIC, is highly correlated with the “integral of deceleration.” From the above, the high correlation between the “integral of deceleration” and the HIC is largely influenced by the secondary impact velocity.

Fig. 8　Comparison of correlation between SIVH and safety evaluation indexes at Ps2 and Ps3

4.3 Proposal of evaluation index and limit value for vehicle body structure

Considering the analysis results and discussion, the “deceleration integral” is considered to be the most suitable crashworthiness evaluation index for the design of the vehicle body structure among the indexes compared this time, because it has a high correlation with passenger injury levels and contains information on seating positions on longitudinal seats, i.e. the time until the secondary impact. When using the “deceleration integral” as an evaluation index for the vehicle body structure, a positive correlation was observed between the “deceleration integral” and the HIC. Therefore, in order to improve crashworthiness, it is desirable for the “deceleration integral” to be low in the vehicle body during a collision. In addition, since it was confirmed that the seating condition of the second seat has a higher risk of head injury than other seating conditions, it is desirable to target this seating condition. From the above, referring to the results of the seating condition of the second seat shown in Fig. 5(a), it is recommended that the limit value for the “deceleration integral” to ensure safety (hereinafter, the vehicle body structure limit value) is set at approximately 4.0 m/s or less, taking into account the injury level limit value of 1000 for the HIC (Fig. 9(a)).

Fig. 9　Comparison of correlation between HIC and safety indexes at Ps2

In Japan, where there are no evaluation standards, the “integrated deceleration value” is optimal. However, with the aim of global standardization, such as the ISO standardization of crash safety evaluation standards, the “120 ms mean deceleration” which is used in the European standard has a good correlation with injury severity, so it was also considered appropriate to select this as the domestic evaluation index. However, the limit value of 5.0 G [1] for vehicle body structures based on this index far exceeds the injury severity limit value of HIC as shown in Fig. 9(b). Therefore, when using it to evaluate vehicle body structures with longitudinal seats in Japan, it is recommended to use a value of approximately 3.0 G or less, taking the injury severity limit value of 1000 into account.

We have presented an evaluation index and its limit value for vehicle body structures that align with the actual situation in Japan, but the severity of injury depends on the interior fittings that are hit in the secondary impact. There are various types of bench-end partitions at the ends of longitudinal seats, and in this report, we used one of the plate-type partitions that has been widely adopted in Japan in recent years. The high correlation between the “integral of deceleration” and injury level is due to the fact that the integral of deceleration has been confirmed to be highly correlated with the secondary impact velocity. Therefore, it is expected that the correlation with injury level will be high even if the partition is of a different plate type. On the other hand, the limit value of the recommended body structure may differ depending on the type of partition (integral of deceleration 4.0 m/s, mean deceleration of 3.0 G over 120 ms). It is expected that the limit value of this evaluation index can be increased by reducing the injury level by changing the partition design. Therefore, the limit value presented in this report is considered to be a strict value (a safe assumption when considering the injury level).

4.4 Future study

The results of this report are based on the assumption that impact deceleration occurs only in the direction of train travel, and that a dummy seated alone on a longitudinal seat experiences a secondary collision with the bench-end partition. In reality, train behavior is not limited to the direction of travel, and is also affected by the pitching of the railway vehicle. However, this situation cannot be reproduced in sled testing, and the accuracy of the analysis cannot be ensured, so this is not included in the analysis conditions. In addition, the analysis accuracy of the behavior and injury level of the dummy model and partition model during a secondary impact was confirmed by comparing with tests using actual in-car equipment such as partitions [10], but the accuracy when applied to an actual accident could not be confirmed. This is because detailed information such as “the location and degree of injury to passengers,” “the in-car equipment causing the injury,” and “the seating position at the time of the accident” in the event of a railway crossing accident has not been made public at present.

Although an analysis was conducted under limited conditions for a situation where multiple dummies were seated on a longitudinal seat, an increase in chest injury values due to secondary collisions with other dummies was observed [14]. However, there are many combinations of the number of dummies seated on the longitudinal seat and the seating positions, and the ES-2 dummy can only evaluate one side of the chest at a time, so the combinations are even more diverse. In addition, since the physique of the ES-2 dummy is larger than that of Japanese people, it is difficult to reproduce a 100% sitting rate in a natural posture on the longitudinal seat, and some ingenuity is required in setting the analysis conditions. For this reason, the evaluation of multiple passengers is considered a future issue.

5. Conclusions

In order to improve the safety of passengers in longitudinal seats during train accidents, we investigated a crashworthiness evaluation method suited to the actual situation in Japan that can be used in the design of car body structures that take crashworthiness into account. We performed crash analyses under various level crossing accident conditions, referring to past accidents that have occurred in Japan. As a result, we proposed the integral value of the car body deceleration as an evaluation index, with the integral time being the time t_imp calculated from the distance between the dummy initial position and the bench-end partition, which showed the highest correlation with the degree of head injury of passengers seated in longitudinal seats. In addition, we proposed a limit value of 4.0 m/s or less for the second seating condition when using this index. We also showed that the 120 ms mean deceleration, which is the evaluation index of the European standard, could be a candidate for the evaluation index, with a view to global commonality, such as the future establishment of ISO crashworthiness evaluation standards. In the future, we plan to use the proposed evaluation index to propose specific car body structures and in-car equipment measures. When determining crashworthiness assessment indexes and their limit values in Japan, it is necessary to hold discussions among relevant government ministries, railway operators, rolling stock manufacturers, and other relevant stakeholders, and we believe that the knowledge gained in this report will be useful at that time.

This report is a reprint of the content published in the Transactions of the Japan Society of Mechanical Engineers in 2021 [15] with some omissions and modifications.

References

• [1]  CEN, EN15227 2020 Railway applications -Crashworthiness requirements for rail vehicles, 2020.
• [2]  Federal Railroad Administration, 49 CFR Part 238 Passenger equipment safety standards, 2011.
• [3]  Nakai, K. and Enami, S., “Experimental validation of effect on reducing the severity of injury to rail passengers occupying longitudinal seats using handrails in the event of collision,” Transactions of the JSME, Vol. 85, No. 878, 2019 (in Japanese), DOI: 10.1299/transjsme.19-00093.
• [4]  Marunaka, T., Taguchi, M., Kishida, K., Kumamoto, H. and Yoshikawa, T., “Rail vehicle design for crashworthiness (1st report, Development of carbody structure for high energy absorption and calculation method for crashworthiness),” Transactions of the Japan Society of Mechanical Engineers, Series A, Vol. 67, No. 664, pp. 72-78, 2001 (in Japanese).
• [5]  Marunaka, T., Taguchi, M., Kimura, T., Kishida, K., Kumamoto, H. and Yoshikawa, T., “Rail vehicle design for crashworthiness (2nd report, Carbody crash test and crash behavior by numerical calculation),” Transactions of the Japan Society of Mechanical Engineers, Series A, Vol. 68, No. 666, pp. 163-168, 2002 (in Japanese).
• [6]  Kimura, S., Mochida, T., Kawasaki, T., Nakamura, H. and Yamaguchi, T., “Evaluation of energy absorption of crashworthy structure for railway's rolling stock (Numerical simulation using damage-mechanics model),” Transactions of the Japan Society of Mechanical Engineers, Series A, Vol. 78, No, 793, pp. 98-112, 2012 (in Japanese).
• [7]  Hamajima, T. and Terumichi, Y., “Train set motion analysis in large displacement in consideration of plastic deformation of a coupler,” Transactions of the Japan Society of Mechanical Engineers, Series C, Vol. 79, No. 806, pp. 497-507, 2013 (in Japanese).
• [8]  Okino, T., Nagata, K., Horikawa, K. and Kobayashi, H., “Evaluation method for crashworthiness of railway vehicles based on the correlation with the degree of passenger injury,” Transactions of the JSME, Vol. 86, No. 881, 2020 (in Japanese), DOI: 10.1299/transjsme.19-00249.
• [9]  Okino, T., Nagata, K., Sato, H., Horikawa, K. and Kobayashi, H., “Evaluation of effect of various collision conditions on safety of occupants in the event of level crossing accident,” Transactions of the JSME, Vol. 85, No. 869, 2019 (in Japanese), DOI: 10.1299/transjsme.18-00270.
• [10]  Nakai, K., Suzuki, D., Enami, S., Okino, T. and Takano, J., “An estimation of the severity of passenger injuries seated on a longitudinal seat in a train accident and measures to mitigate it by using numerical simulation,” RTRI report, Vol. 33, No. 1, pp. 29-34, 2019 (in Japanese).
• [11]  Okino, T., Yamamoto, M., Takano, J. and Ujita, Y., “Crashworthiness evaluation for train carbody in the event of level crossing accident by driver's injury indexes,” Proceedings of the 19th Jointed Railway Technology Symposium 2012, No. 12-79, pp. 557-560, 2012 (in Japanese).
• [12]  Ministry of Land, Infrastructure, Transport and Tourism, Announcement details appendix 24: Technical standard for the occupant protection device in the event of a lateral collision, The new safety standard of the road transport vehicle, pp. 769-811, 2009 (in Japanese).
• [13]  Nakai, K., Suzuki, D., Enami, S., Okino, T., Takano, J. and Palacin, R., “An Estimation of behaviour and severity of injury to rail passengers occupying longitudinal seats in the event of collision,” Proceedings of IRCOBI Conference 2015, pp. 315-326, 2015.
• [14]  Omino, K., Nakai, K., Shiroto, H. and Suzuki, D., “Simulation of behavior of passengers aboard commuter train in the event of level crossing accident,” RTRI report, Vol. 26, No. 1, pp. 21-26, 2012 (in Japanese).
• [15]  Nakai, K., Enami, S., Okino, T. and Nagata, K., “Evaluation method for crashworthiness of railway vehicles based on the correlation with the injury severity of passengers occupying longitudinal seats,” Transactions of the JSME, Vol. 87, No. 903, 2021 (in Japanese), DOI: 10.1299/transjsme.21-00181.

Authors

	Kazuma NAKAI, Dr.Eng. Senior Chief Researcher, Head of Ergonomics Laboratory, Human Science Division Research Areas: Interior Passive Safety, Human Dynamics, Ergonomics
	Tomohiro OKINO, Ph.D. Senior Chief Researcher, Head of Vehicle & Bogie Parts Strength Laboratory, Vehicle Technology Division (Former) Research Areas: Crashworthiness, Carbody Strength
	Shota ENAMI Senior Researcher, Ergonomics Laboratory, Human Science Division Research Areas: Interior Passive Safety, Ergonomics
	Keisuke NAGATA Senior Researcher, Vehicle & Bogie Parts Strength Laboratory, Vehicle Technology Division Research Areas: Crashworthiness, Carbody Strength