Influence of Fixing Feet in Rear-End Crashes

Abstract

Considering occupants’ safety in highly automated vehicles (HAVs) is one of the future trends in mobility. Passengers will not deal with driving the car anymore, as a result of the automated driving system (ADS) application. Seat backs can be reclined by occupants during long trips to provide a comfortable posture. Vehicle collisions will still be possible, regardless of ADSs being used in HAVs, as many cars will still be driven manually. This study follows two objectives. The first is to seek the influence of restricting the body’s feet on a body’s kinematic. The second aim is to investigate and compare the model’s body part injury criteria for upright and reclined seating postures with and without feet fixation. Finite element simulations are performed with a 50th percentile Virthuman model placed on a deformable seat and fastened to the seat by a three-point belt. Outcomes show that, if a body is in an upright posture, fixing models’ feet to the interior causes most injury criteria of body parts to be improved and will lead to safer conditions in comparison with accidents in which the model’s feet are free to move. The exception is the model’s tibiae, whose injury criterion was worsened. On the other hand, if the model is in the reclined position, fixing the model’s feet could improve the neck, abdomen, and knee injury criteria. Meanwhile, femur and tibiae injury criteria noticeably worsen.

Keywords

safety occupant protection safety performance and analysis crash analysis modeling and forecasting

In the future, vehicles’ occupants will be merely passengers and they will not need to be involved in driving their cars. An automated driving system (ADS) will take full responsibility for controlling a vehicle’s driving during trips ( 1 ). ADS can potentially reduce the chance of cars colliding. It is possible to reduce rear-end striking crashes and rear-end striking crashes with injuries by 50% and 56%, respectively, by using both forward collision warning (FCW) and autonomous emergency braking (AEB) ( 2 ). But autonomous vehicles are in the phase of research and development and traffic accident records for them are limited ( 3 ). On the other hand, there are more conventional vehicles on streets and motorways which will be driven by drivers and do not use this advanced equipment. It has to be also added that the conventional vehicles’ drivers need time to become accustomed with autonomous vehicles reactions ( 3 ). They drive either too close to an autonomous vehicle, or with unsafe speed. Therefore, it is quite possible to expect crashes between normal cars and highly autonomous vehicles. Since 2014, the number of connected and autonomous vehicles (CAV) has been continuously increasing ( 4 ). Statistics show that the rear-end crash for autonomous vehicles is the predominant collision type (57.5%) among different collision types ( 4 ). Another survey’s outcome reveals that 60% of crashes reported related to autonomous vehicles in California in 2019 were rear-end ( 5 ). This is almost five times more than what was recorded for head-on collisions.

The interior of a highly automated vehicle (HAV) faces revolutionary changes because of the absence of some controlling devices such as the steering wheel and gear lever. These components will no longer exist in fifth-generation vehicles, as a result of using ADS. In this regard, more space will be available for passengers in the coming future. HAV occupants can opt for one of several comfortable seating positions as a result of being able to adjust their seats either by reclining their back or rotating them. Now, an important question for designing an HAV interior is: Are fully autonomous vehicles safe enough for their occupants?

The new interior designs for fully automated vehicles are not obliged to be in an upright position. It is a traditional seat posture in which passengers sit in a vehicle’s seat as in the standard driving style. In an HAV interior, all passengers can recline their seats’ backs to get a more comfortable posture that is the desired posture for occupants, especially during long trips. This study considers two well-known positions are: (a) a seat that has an upright back and (b) a seat with a reclining back ( 6 ).

A virtual human body model, Virthuman, was chosen to represent a human body in the present study. The Virthuman is a hybrid scalable virtual model. Its skeleton is formed as a multibody system (MBS) structure. The model’s skin surface is made of rigid surface parts that are named super elements and are interconnected by strips of elements with no mechanical response. Therefore, it is suitable for fast calculations. Springs and dampers connect the MBS structure to the model’s skin. The springs and dampers are tuned to assure biofidelic deformation in impact tests ( 7 – 11 ). For this reason, an extensive validation process has been performed with the Virthuman model. Validation tests, including impacts with individual body parts as well as whole-body tests, have been reconstructed. These are based mostly on available experimental data from post-mortem human subjects (PMHS) mechanical responses ( 7 ). In the current study, it is important to know the biofidelity of the model with regard to the lower mechanical loadings. For instance, biofidelity of the Virthuman under low-g impacts has been proven during the validation of the neck which is based on data with volunteers as well as during the low-g sled test ( 7 , 10 ).

The Virthuman can represent a wide spectrum of occupants as far as their gender, height, weight, and age are concerned ( 12 ). Thus, the model can be simply adapted to any initial position, for example, upright and reclined, in a virtual environment. Injuries are measured for different parts of the body, that is, head, neck, thorax, pelvis, femur, knee, and tibiae. For better visualization, these injury criteria will be presented on each body’s region using colored segments concerning simulation time ( 13 ).

Autonomous vehicle usage is not still so common among the public. There are not too many traffic accident records/safety simulations available for them. The present study’s motivation is to investigate the possibility of considering a conceptual foot support design that might help increase occupants’ safety in rear-end crashes. Therefore, four simulations in two groups are planned to seek the influence of human body pre-positioning on its kinematic and injury criteria. The body is located in a deformable seat which is made from polyurethane foam. Two dominant variables are considered for the body’s feet: free and supported. In the first case, the feet are not in contact with any of the interior’s points/elements and can move freely during the simulation. For the second situation, the model’s feet are fixed to the interior’s floor with an unbreakable connection.

Method

The PAM-Crash module in the Virtual Performance Simulation (VPS) environment was used for generating models and running simulations. The virtual human model (see Figure 1) which was used in this survey represents a 30-year-old male with a height of 176 cm and a weight of 78 kg (50th percentile male).

Figure 1.

Virthuman model in upright position.

The body is located in a deformable seat. The seat’s geometry is based on the driver’s seat of the Volvo XC70 Station Wagon (see Figure 2) ( 14 ). The seat’s total weight is slightly more than 28 kg. According to the survey’s goal, two positions for the seat’s back are considered. One is called upright, where the seat’s back is positioned upright and its angle with the model’s vertical axis (z) is 15.43 degrees (see Figure 2). The second case is a reclined seat, in which the angle between the seat’s back and the model’s vertical axis is 45.70 degrees. The seat is made from polyurethane foam which is widely used in the automobile industry.

Figure 2.

Deformable seat: upright seat (left) and reclined seat (right).

The foam is modelled by using general nonlinear strain rate foam material in a VPS environment. It is a highly compressible non-linear elastic foam with strain-rate dependency and optional energy absorption (hysteresis). The material model is validated for crash simulations ( 15 ). For describing the foam’s behavior, two compressions and tension stress curves versus engineering strain have to be determined in VPS (see Figure 3). The foam’s Young modulus and unloading energy dissipation factors are 0.02 Gpa and 0.25, respectively.

Figure 3.

Polyurethane foam material properties.

As the seat model is a deformable one, a pre-simulation (free-fall) had to be done first to make initial deformations to the seat. For this reason, the body model was released slightly above the seat from rest. Pure gravitational acceleration was applied to whole model nodes (seat, interior, and body) along the z-axis. For simplicity reasons, the seat does not include any internal components such as steel bars and springs. Instead of these elements, some nodes’ motion is fixed. These nodes were chosen under the seat’s cushion, back, headrest, and armrest (see Figure 4).

Figure 4.

Fixed nodes on the seat.

A pre-simulation was run for 3 s which was quite enough for the model to reach a stable position. The body’s surface penetrated the seat surface and deformed it. The deformed seats are illustrated in Figure 5. As can be seen for the upright seat, the majority of the deformation occurred in the seat’s cushion. However, both the seat’s back and cushion were deformed in the reclined seat.

Figure 5.

Deformed seat: upright seat (left) and reclined seat (right).

For fastening the body to the seat, an integrated continuous three-point belt was used (see Figure 6). The seatbelt was generated by the Safe Tools in VPS and was 1.2 mm thick with a width of 40 mm. The belt’s key parts are connecting bars, and shoulder and lap straps. The seatbelt’s strap is a two-layered anisotropic membrane. Its material parameters correspond to the common seatbelt material used in current passenger vehicles. The seatbelt materials’ loading and unloading curves versus engineering strain (see Figure 7) was adopted for numerical simulations such as in the sled test from mechanical response point of view ( 10 , 16 ). The shoulder belt starts from a D-ring, then it passes over the body’s trunk diagonally from the upper left to bottom right (from D-ring to buckle). Meanwhile, the lap belt is located over the model’s belly from right to left (from buckle to anchor). Non-linear bars are used to connect the free ends of the straps to the external parts of a seatbelt (retractor, buckle, and anchor) (see Figure 6). The nonlinear bar segment starts from the retractor, passes through the D-ring, and ends at the upper part of the shoulder strap. Another nonlinear bar part connects the shoulder and lap straps via the buckle. Last, but not least, is a bar division that connects the other side of the lap belt to the anchor.

Figure 6.

Three-point seatbelt.

Figure 7.

Seatbelt’s components’ material loading curves versus engineering strain. left) bar, right) strap.

The last part of the model is the interior (see Figure 1). It was modelled by simple shell elements with 50 mm thickness.

Crash Simulation

A 30 km/h crash acceleration pulse is used for a frontal crash in this study (see Figure 8) ( 16 ). The pulse is applied horizontally to the model’s centre node along the x-axis toward the minus direction to simulate a rear-end collision.

Figure 8.

Crash acceleration pulse.

Four simulations were run in two groups: upright seat back and reclined sat back. In each group, there were also two subgroups. In one instance, the body’s feet were fixed to the interior (with foot support). In other cases, the feet are free to move and no fixation is considered for them (with no foot support). An unbreakable tied link was used for modelling the feet’s support to the interior (see Figure 9). The Virthuman was generated based on the MBS, and external boundary conditions do not apply to the MBS structure in VPS. Therefore, tied supports were chosen for fixing the feet to the interior’s bottom.

Figure 9.

Foot support.

Part Injury Assessment

An algorithm was developed to assess the injury criteria for different body parts on the Virthuman model ( 17 ). The program uses the simulation’s outcomes to generate 779 time-dependent curves, for example, accelerations, forces, and displacements of certain points and parts of the Virthuman model. A list of injury criteria that are evaluated for the Virthuman model is given in Vychytil and Špirk ( 18 ). For instance, the head’s centre of gravity is picked to store the acceleration’s component and its absolute value to be considered for computing the head injury criterion (HIC). For computing injury criteria of the neck, its joints’ flexion moment, tensile, and shear forces play important roles. The program measures the deflection of the model’s thorax segments and also computes viscous criteria (VC) during the simulation. Based on the last-mentioned values, it returns the level of injury for the model’s chest. Compression forces on abdominal parts are used to determine the injury level in the model’s abdomen. Injuries for pelvis segments are evaluated based on pelvic forces. For the femur, knee, and tibiae, as well, forces and moments of the breakable joints are considered for the assessment of these parts’ injury level.

A criterion of injury for each model part (e.g., head, neck) is determined according to its relevant time-dependent curves. The worst degree of injury of a particular segment in the time interval from the beginning to the current time step is assigned as its injury criteria ( 17 ). On this account, the injury’s value for each part is read from the last animation step.

It is worth mentioning that the body part injury values for the Virthuman model are age-dependent, but the body’s height, mass, and gender are not considered for these ranges. The injury criteria metrics are available in the literature for a 6-year-old child, 20-year-old, and 100-year-old adult for each injury criteria ( 19 – 21 ). For the other ages, the program uses linear interpolation to compute the metrics. For instance, the same load state causes worse injuries in children than in adults. Injury criteria are returned by four basic levels (color coded) according to the EuroNCAP consumer rating ( 13 ). Either a small degree of injury or none is given as “Good.” An injury level can also be “Acceptable” or “Marginal.” However, fatal injuries are shown by the “Poor” level and cause a very serious degree of injury. As an example of the HIC, if the HIC36 is lower than 650, a “Good” level is assigned to the head and it is represented by green on the model in the software’s viewer. When HIC36 is more than 1,000, the program returns “Poor” and colors the head in the software’s viewer red.

A point that must be highlighted is that each body part includes some segments, that is, the Virthuman’s thorax contains 46 segments. The injury assessment algorithm considered the worst injury criteria of these segments in each time step to demonstrate the injury criteria of the part. For instance, in a time step, the injury criteria of a single segment in the thorax are “Acceptable.” The rest of the segments in this body part have “Good” injury criteria. Therefore, the general assessment for the thorax injury criteria is “Acceptable.”

Discussion

Kinematic

Bodies’ kinematics are recorded from the simulations for four nodes on the body: head centre, H-point, knees, and feet. These nodes’ motion trajectories for two postures with and without foot supports are shown in Figure 10. It is observed—in general—that a body faces a backward motion in its sagittal plane when a rear-end crash happens. The body moves up from the seat cushion and the contact between the body’s pelvis and the seat is cut. While the crash is happening, the body’s back and head hit the seat back and headrest, respectively. In rebound, the body moves forward. The shoulder belt stops the trunk’s motion and the lap belt does not allow the pelvis to move forward. This causes the body to fall on the seat and hit the cushion. A slight forward movement is seen in the body.

Figure 10.

Body’s kinematic: Upright posture (left) Reclined posture (right).

When an upright posture is considered, fixing the feet to the interior does not make any obvious difference to the body’s head and pelvis trajectory curves (see Figure 10). When the feet are fixed, the knees experience diagonal movement, backward and forward in the body’s sagittal plane. But in the test case in which feet can have free movement, the knees at first move backwards in a plane that is parallel to the interior’s floor. Then in the rebound, the knees have a forward upward motion. The last-mentioned knees’ movement causes them to hit the interior’s front. The movement of points located on the body’s ankles is quite similar to the knees’ motion.

The second case study inspected a body’s behavior when it is located on a reclined seat. The model’s head motion was captured from a node located at the head’s centre of gravity and elaborated that, if the feet are fixed, the head at first hits the headrest and takes an upward motion toward the headrest. Then, during the rebound, it moves forward almost parallel to the interior’s floor plane. If the feet are not fixed, the model’s head maintains its motion back and will not experience any forward motion in its rebound.

A remarkable difference is seen between the body’s pelvis motions in the two test cases (with and without foot support). In the case where the feet are fixed to the interior, during the crash, the pelvis moves backwards and then undergoes diagonal forward movement. Then it returns almost to its initial position. But if the feet are free to move, the pelvis moves backwards further and, by the end of the crash test, it is located above the cushion. The knees’ motion in this test case is different for the model with and without foot support. Whenever the feet’s motion is restricted, the knees do not have too much movement. The body’s tibiae and femurs just bend around the knees’ joint and cause the knees to hit the front of the cushion. Somehow, at this moment, the tibiae and femurs are aligned. The knees return to their initial position and the leg segments are more or less located in their initial positions as well. On the other hand, when no restriction is applied to the feet, the knees move backwards at first. Because the body starts to move up from the seat cushion, the knees also move up. Afterwards, they move down, but quite far from their initial position. Here, it is possible to see the ankles’ trajectory and the motion of the knees have the same tendency.

Body Injury Levels

Injury criteria for the body parts in a rear-end crash test are determined based on injury evaluations according to the EuroNCAP consumer rating in four colors on different parts of the body: head, neck, thorax, abdomen, pelvis, femurs, knees, and tibiae ( 13 ). Green is used for a low degree of injury of a body part. It is interpreted as a “Good” level. The algorithm which computes injury criteria for body parts on the Virthuman also uses two other color codes to demonstrate a body part injury criteria on the model. If an “Acceptable” injury criterion is assessed for a body part, it is shown in yellow. On the other hand, a body part whose injury criterion is “Marginal” is shown in brown. The last, but not least, injury criterion is “Poor” which means fatal injuries can happen on this body part. Therefore, red is chosen to show these body parts.

Head Injury

From Figure 11, it is seen that the index of injury for the model’s head is “Good” in two crash tests with and without foot support. If a model’s head hits anywhere, a jump occurs in the head’s acceleration. But here, a deformable headrest is used which can absorb crash energy. Therefore, the HIC was lower than 650 and the injury criteria assessment algorithm assigned the safest level, green, to the model’s head. The value of the HIC for the model in the reclined posture is lower than the model’s HIC in the upright posture.

Figure 11.

Head injury criterion (HIC): Upright posture (left) Reclined posture (right).

Neck Injury

Considering the body’s neck, while the upright posture is taken into account, the forces and moments in the upper joints of the neck do not exceed the limitations defined in Mecas ( 17 ). Therefore, the neck’s injury criteria remain “Good” during the simulation (see Figure 12). For the reclined seat, if the model’s feet are fixed to the interior’s floor, the neck injury criteria are “Marginal” as a result of an increase in the neck’s joint extension moment (see Figure 12). On the other hand, for the test where the feet are free to move, the absolute value of the neck’s extension moment is slightly higher than the test case without foot support. This leads to the “Poor” injury criterion.

Figure 12.

Non-dimensional neck extension moment versus simulation time: upright posture (top) and reclined posture (bottom).

Thorax Injury

The body undergoes a backward motion in its sagittal plane because of the rear-end crash test, in the present study. The model’s back hits the seat’s back first and the rear segments of the body’s thorax are pressed inward. As result, deflection of the joints that are located beyond the thorax’s segments is increased. Of course, it is not too much. Meanwhile, the shoulder belt restrains the model’s forward motion. It presses the thorax’s frontal and side segments. Figures 13 and 14, respectively, illustrate the frontal and side deflection of a segment in a body’s thorax that has the highest values. In comparison with the thorax injury criteria given in Mecas, the model’s thorax injury criterion is “Acceptable” for upright and reclined seating postures with and without foot support ( 17 ). This is because either the thorax’s front or side deflection exceeds the “Acceptable” criterion threshold.

Figure 13.

Non-dimensional thorax frontal deflection versus simulation time: upright posture (top) and reclined posture (bottom).

Figure 14.

Non-dimensional thorax side deflection versus simulation time: upright posture (top) and reclined posture (bottom).

Abdomen Injury

The model’s abdomen is the other body part whose injury criterion will be discussed. Here, the compression forces of the abdominal segments are taken into account for determining the model’s abdomen. A segment whose compression force versus simulation time was chosen and the force’s changes against time are given in Figure 15. The two curves that belong to seating postures and foot support type show the same tendency. Plots can be cut into three independent sections. In the first section, no obvious changes are seen for the abdominal force. This is because of the body’s inertia. The seat moves forward because of the rear-end crash but the body remains in its position (see H-point curves in Figure 10). Then, a raise is seen in the force in the plot. Here is the point where the model wants to move up but the lap belt prevents the motion. Kinematic curves of the H-point illustrate the body’s motion in Figure 10. The slope of the motion’s curves for a body with upright posture is steeper than the reclined posture. This means the lap belt tightly pushes back the body to its initial position. In this regard, the pick point of the curve—in absolute value —for the upright posture is higher than the reclined posture. The last section on the abdominal force plot versus time concerns when the body falls onto the seat. The contact between the lap belt and body will not exist anymore as far as no retractor is acting on the lap belt. Therefore, the compression force on the abdomen’s segments will be decreased.

Figure 15.

Non-dimensional abdomen compression versus simulation time: upright posture (top) and reclined posture (bottom).

Pubic Injury

Pubic forces are key points for injury criteria assessment on the Virthuman pelvis. In Figure 16, pubic forces and their injury criteria are shown for the body’s pelvis and are illustrated in two postures. Meanwhile, the effect of having foot supports is also demonstrated. The upright posture without foot support is the only case that has a lower injury criterion, “Acceptable,” in comparison with other tests in the present study. For the test case, where the model’s feet are free, the model’s legs move upward and the femurs are bent backwards to the belly around the h-point. The pubic force exceeds the injury criterion that is given for the pelvis in Mecas ( 17 ). With returns to the simulations’ outcomes, the injury criterion for the model’s pelvis in the reclined posture is better in comparison with the upright posture.

Figure 16.

Non-dimensional pubic force versus simulation time: upright posture (top) and reclined posture (bottom).

Femur Injury

The next issue which is evaluated for Virthuman’s injuries is the femur injury criteria, in which compression force and moment of femur parts must be considered. From Figure 10, it can be seen the body moves up from and falls onto the seat cushion because of the rear-end crash. This motion changes compression forces. The compression force for a femur’s segment with the highest peak is shown in Figure 17. The model’s femoral injury criterion in the upright posture is “Acceptable.” In the reclined posture, a sharp jump is seen in the femoral compression force. A numerical error caused the femur injury criterion for the case with foot support to be determined as “Marginal.”

Figure 17.

Non-dimensional femur compression force versus simulation time: upright posture (top) and reclined posture (bottom).

Knee Injury

The knee joint connects the femur to the tibia and its moment is used for evaluating the knee’s injury criterion on the body model. Whenever the model’s heels are fixed to the interior’s floor, movement of the femur and tibia causes a slight rotation in the joint. Knee moments for a joint —that includes the highest peak value —against simulation time are illustrated in Figure 18. The plots in Figure 18 show that having no foot supports increases the knee injury criterion. For the upright posture, the knee injury criterion is “Poor” if the feet are not fixed, while fixing the feet improves the knee injury criterion to “Good.” On the other hand, considering the reclined posture, in the case of free feet, the knee injury criterion is “Acceptable.” With fixed feet, the knee injury criterion changes to “Good.”

Figure 18.

Non-dimensional knee moment versus simulation time: upright posture (top) and reclined posture (bottom).

Tibia Injury

The compression force for a tibia’s segment with the highest peak is shown in Figure 19 with regard to the two seating postures and foot support conditions. When the feet are fixed to the interior’s floor during the rear-end crash test, the body’s motion toward the seat back forces the tibiae to rotate clockwise around the ankle joints.

Figure 19.

Non-dimensional tibia compression force versus simulation time: upright posture (top) and reclined posture (bottom).

Fixed heels restrict ankle translational motions. Therefore, tibiae compression forces are increased. If the model’s heels are fixed to the interior’s floor, the injury criteria of the model’s tibia for the upright and reclined postures are “Acceptable” and “Poor,” respectively. On the other hand, when the feet are free to move, the injury criteria for the upright posture is “Good” because the tibia’s compression force plot against simulation time does not exceed the green band. In addition, for reclined postures, the tibia injury criterion is “Acceptable” as far as the plot crossed the yellow band.

Dominant Injury Occurrence Time

The occurrence time of the final injury level—dominant one—per body part is illustrated in Figures 20 and 21 for the upright and reclined posture tests respectively. The dashed line represents the simulation timeline. The body parts which have a “Good” level from an injury criterion point of view are not shown in Figures 20 and 21. They just show the injury occurrence time for the parts which have other level injuries. The colored boxes in the plot demonstrate three pieces of information about its relevant body section: the injury sequence for parts (roman numeral), body part name and the first time that the dominant injury started. For instance, in Figure 20 when the model’s feet are free, permanent injury first happens to the model’s thorax at 61 ms. After that its pelvis will be at the “Acceptable” injury level at 68 ms. The last body section that reaches its dominant injury is the model’s knees, which reach this level from 74 ms.

Figure 20.

Dominant injury criteria occurrence time for upright posture: without support (top) and with support (bottom).

Figure 21.

Dominant injury criteria occurrence time for reclined posture: without support (top) and with support (bottom).

According to Figure 20, it is seen that fixing a model’s feet during a crash could increase the number of body parts with a “Good” injury level. The second point is the injury levels are reduced from a quality point of view, somehow, such that no body part with a “Poor” injury level is seen in the sequence diagram. Fixing the feet could also postpone the final injury level for the model’s thorax and femurs. This time delay is not that great for the thorax; however, it is remarkable for the femurs. Abdominal injury level is reduced from “Poor” to “Marginal” if the model’s feet are fixed to the interior. However, it cannot cause a time delay in the occurrence of injury and it happens more or less at the same time. Last, but not least, is an injury to the model’s tibiae that happens because of the model’s feet being fixed. The tibia injury level changes from “Good” to “Acceptable.”

The influence of fixing the model’s feet to the interior on a model’s part injury level from an injury level sequence point of view is given in Figure 21 for the test case in which the model is in the reclined posture. Here, it is also possible to see that fixing has a quite similar effect on the body parts injury level. It reduces the level of injury to the upper parts of the body. For instance, the neck injury level, which is “Poor” if the motion of the feet is not restricted, changes to “Marginal.” However, the time in which neck injury occurs does not change too much. There will be no abdominal and knee injury for the model if the feet are fixed. On the other hand, the “Marginal” injury level detected for the femurs and tibiae move to the “Poor” injury level even sooner than the test case where the model has free foot motion. The other point is the model’s seating posture causes an almost 20 ms time delay in reaching an “Acceptable” injury level for the model’s thorax.

Conclusion

The current study investigated the influence of restricting the body’s feet on a body’s kinematic as well as the injury criteria of body parts in rear-end crashes. In this regard, a 50th percentile Virthuman male model was chosen for running finite element simulations. Two seating postures, upright and reclined, are considered here while the model’s feet can be affixed to the interior or are free to move. Therefore, four tests were planned to simulate the crash. The body was placed in a deformable seat which is made from polyurethane foam, and a three-point seat belt with no initial pre-tensioner was used to fasten the body to the seat.

Concerning the simulations results, it was shown that, if the model’s feet are fixed to the interior during the crash test in the upright posture, most of the body parts injury criteria improve and there will be safer conditions in comparison with when the model’s feet are free to move. The body’s head and neck injury criteria levels do not change. But fixing the feet increases tibiae compression force and causes this body part’s injury criterion to worsen.

For the other crash tests scenarios when the Virthuman is in the reclined posture, fixing the model’s feet could improve the neck, abdomen, and knee injury criteria. Meanwhile, the model’s injury criteria for the femurs and tibiae noticeably change and worsen. In addition, the model’s head and pelvis do not see any changes from an injury criteria point of view.

Returning to the present survey’s outcomes, integrating foot support into cars’ next generation (autonomous vehicles) can improve occupants’ safety. For simulation purposes, here, the model’s feet are fixed by using tides. But, in reality, passengers’ foot supports will be a complex mechanism that is in connection with the ADS. When the vehicle is driven normally, the foot support is not active and does not disturb a passenger’s feet’s free movement. Whenever the ADS detects a crash and there are no possibilities to avoid the collision (vehicle maneuver, braking, etc.) the foot support mechanism is activated and grabs (affixes) an occupant’s tibiae (mainly the tibiae’s lower segments). The inner side of the supports is covered by soft material, for example, polyurethane foam, to avoid injuries on passengers, tibiae and feet. It will help a lot to reduce the injury levels on a model’s tibiae in the crash. A question is still open for using this foot support for autonomous vehicle occupants: a passenger’s height. For instance, a toddler is not tall enough and their tibiae are not well located in this safety device. Therefore, the crashworthiness of this device has to be simulated for various passenger heights.

Footnotes

Author Contributions

The authors confirm contribution to the paper as follows: study conception and design: A. Talimian, J. Vychytil, Luděk Hynčík; data collection: A. Talimian, J. Vychytil, Luděk Hynčík; analysis and interpretation of results: A. Talimian, J. Vychytil, Luděk Hynčík; draft manuscript preparation: A. Talimian, J. Vychytil, Luděk Hynčík. All authors reviewed the results and approved the final version of the manuscript.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the European Regional Development Fund-Project “Application of Modern Technologies in Medicine and Industry” (No. CZ.02.1.01/0.0/0.0/17_048/0007280).

ORCID iDs

Abbas Talimian

Jan Vychytil

Luděk Hynčík

References

NHTSA. Automated Vehicles for Safety. 2017. https://www.nhtsa.gov/technology-innovation/automated-vehicles-safety#topic-road-self-driving.

Cicchino

J. B.

Effectiveness of Forward Collision Warning and Autonomous Emergency Braking Systems in Reducing Front-to-Rear Crash Rates. Accident Analysis & Prevention, Vol. 99, 2017, pp. 142–152.

Petrović

Đ.

Mijailović

Pešić

Traffic Accidents With Autonomous Vehicles: Type of Collisions, Manoeuvres and Errors of Conventional Vehicles’ Drivers. Transportation Research Procedia, Vol. 45, 2020, pp. 161–168.

Ding

Wang

Statistical Analysis of the Patterns and Characteristics of Connected and Autonomous Vehicle Involved Crashes. Journal of Safety Research, Vol. 71, 2019, pp. 41–47.

Fehr

Kempter

Staudenmeyer

Kleinbach

Influence of Prepositioning of the VIVA Model in the Simulation of Rear-End Impact Scenarios. Proc., 8th International Symposium: Human Modeling and Simulation in Automotive Engineering, Wiesbaden, Germany, 2020.

Hasija

Kelkar

Takhounts

E. G.

, eds. Simulation Assessment of Injury Trends for 50th Percentile Males Using Potential Seating Configurations of Future Automated Driving System (ADS) Equipped Vehicles. Proc., 26th International Technical Conference on the Enhanced Safety of Vehicles (ESV): Technology: Enabling a Safer Tomorrow, Eindhoven, Netherlands, National Highway Traffic Safety Administration, Washington, D.C., 2019.

Vychytil

Manas

Cechova

Spirk

Hyncik

Kovar

Scalable Multi-Purpose Virtual Human Model for Future Safety Assessment. SAE Technical Paper 2014-01-0534. SAE 2014 World Congress & Exhibition, 2014.

Lindtstedt

Vychytil

Dziewonski

Hyncik

Numerical Tests of the Virtual Human Model Response Under Dynamic Load Conditions Defined in Federal Aviation Regulation Part 23.562 and 25.562–Preliminary Study. Archive of Mechanical Engineering, Vol. 63, No. 4, 2016, pp. 511–530.

Mecas

E. G.

Virthuman User’s Manual (VPS Explicit MBS Model, Model Version 1.5). Mecas ESI, Pilsen, Czech Republic, 2020.

10.

Špička

Čechová

Špirk

Hynčík

Vychytil

Kovář

Comparison of the Virthuman Model and Hybrid III Dummy at Low-Speed Frontal Impact. Proc., 38th FISITA World Congress, Prague, Czech Republic, September 13–17, 2021.

11.

Talimian

Vychytil

Numerical Study of Frontal Collision Effects on an Occupant’s Safety, in Autonomous Vehicles, With Non-Standard Seating Configurations. Acta Polytechnica Hungarica, Vol. 18, No. 6, 2021, pp. 127–140.

12.

Hyncik

Cechova

Kovar

Blaha

On Scaling Virtual Human Models. SAE Technical Paper 2013-01-0074, Asia Pacific Automotive Engineering Conference, 2013.

13.

EuroNCAP. Assessment Protocol – Adult Occupant Protection. June 2020. https://cdn.euroncap.com/media/58227/euro-ncap-assessment-protocol-aop-v912.pdf

14.

Ozmumcu

Volvo xc7 Station Wagon Seat. 2012. https://grabcad.com/library/automobile-seats.

15.

Špirk

Křížek

Jeníček

Š.

, eds. Polyurethane Foam Behaviour During Impact. Proc., MATEC Web of Conferences. EDP Sciences, Les Ulis Cedex A, France, 2018.

16.

Vezin

Bruyere-Garnier

Bermond

Verriest

J. P.

Comparison of Hybrid III, Thor-α and PMHS Response in Frontal Sled Tests. Stapp Car Crash Journal, Vol. 46, 2002. https://doi.org/10.4271/2002-22-0001.

17.

Mecas

E. G.

Virthuman Postprocessing Manual (VPS Explicit MBS Model, Model Version 1.1). Mecas ESI, Pilsen, Czech Republic, 2016.

18.

Vychytil

Špirk

Numerical Analysis of Passenger Kinematics and Injury Risks During a Railway Vehicle Collision: The Effect of Safety Belts. Applied and Computational Mechanics, Vol. 14, No. 1, 2020, pp. 1–16.

19.

Mertz

H. J.

Patrick

L. M.

Strength and Response of the Human Neck. SAE Transactions, Vol. 80, 1971, pp. 2903–2928.

20.

Kleinberger

Sun

Eppinger

Kuppa

Saul

Development of Improved Injury Criteria for the Assessment of Advanced Automotive Restraint Systems. NHTSA Docket 98-4405-9. US Department of Transportation, Washington, D.C., 1998, pp. 12–17.

21.

Forman

J. L.

de Dios

Symeonidis

Duart

Kerrigan

J. R.

Salzar

R. S.

Balasubramanian

Segui-Gomez

Kent

R. W.

, eds. Fracture Tolerance Related to Skeletal Development and Aging Throughout Life: 3-Point Bending of Human Femurs. Proc., International Conference on the Biomechanics of Impact (IRCOBI), Dublin, Ireland, 2012, pp. 529–539.