Printer Friendly

Influence of pre-impact pedestrian posture on lower extremity kinematics in vehicle collisions.


Lower extremities are the most frequently injured body regions in vehicle-to-pedestrian collisions and such injuries usually lead to long-term loss of health or permanent disability. However, influence of pre-impact posture on the resultant impact response has not been understood well. This study aims to investigate the effects of preimpact pedestrian posture on the loading and the kinematics of the lower extremity when struck laterally by vehicle. THUMS pedestrian model was modified to consider both standing and mid-stance walking postures. Impact simulations were conducted under three severities, including 25, 33 and 40 kph impact for both postures. Global kinematics of pedestrian was studied. Rotation of the knee joint about the three axes was calculated and pelvic translational and rotational motions were analyzed. Pedestrian in walking posture exhibited larger knee bending angle (40% for ipsilateral knee joint) and pelvic rotation angle (27.5% for Z-direction pelvis rotation angle) with less constraint due to isolated single-leg interaction with vehicle and nonplanar characteristic from the leg swing. The walking posture increased the injury risk of soft connection tissue about 20-30% and reduced the internal force in bone structure about 25% regardless of impact severity. Two-leg interaction, inertial effect, anatomical features of the knee and pelvis exhibited a coupled influence on lower extremity kinematics. Injury predictors such as tibia stress and collateral ligament stretching ratio was found to be associated with kinematics. The trade-off of injury risks induced by kinematics with different pre-impact postures is a challenge for vehicle front-end structure design. Further research efforts are necessary to include more loading scenarios and to quantify the lower extremity injury risk in detail.

CITATION: Tang, J., Zhou, Q., Nie, B., Yasuki, T. et al., "Influence of Pre-impact Pedestrian Posture on Lower Extremity Kinematics in Vehicle Collisions," SAE Int. J. Trans. Safety 4(2):2016, doi:10.4271/2016-01-1507.


Lower extremity is one of the most frequently injured body regions in vehicle-to-pedestrian collisions and lower extremity injuries comprise almost 32.6% of AIS2+ injuries [1]. Lower extremity injuries usually take long time to recover [2]. Pedestrian Crash Data Study (PCDS) had documented 521 pedestrian accidents at six sites in United States during the period from July 1994 to December 1998. Data in PCDS pointed out that 68% of the involved vehicles were passenger cars. At the instant prior to crash, 55% of pedestrians were in a walking state and 24% were stationary or not in an obvious motion. 63% of the pedestrians had two legs apart from each other in a lateral sight [1].

To investigate injury mechanisms of pedestrian lower extremities, many cadaveric tests were conducted, using full cadavers or cadaveric parts [1][4][5][6][7]. Full-scale pedestrian cadaveric tests with vehicle front-end revealed that bone fracture and ligament rupture were main injury patterns in lower extremity injuries [3][10]. In full cadaveric tests, pedestrians were always positioned in mid-stance walking posture with almost no overlap between the two legs. Thus the experiments were not capable to represent those nearly 40% crashes in which the two legs interacted. Pedestrian gait had been taken into consideration in accident reconstruction, while has not received enough attention in their influence on kinematics and injury [8][9]. Previous studies of reconstruction focused on relation between head impact location and pedestrian initial postures. Whether pedestrian posture could affect load path and distribution of local deformation in the lower extremities, causing differences in injury patterns and locations remains unclear. Recently the Crash Injury Research and Engineering Network (CIREN) recorded detailed whole body injury data in 67 pedestrian crashes. In-depth statistical analysis demonstrated different frequency of main injury patterns comparing to what cadaveric tests presented [10]. In real-world accidents the pedestrian would not always in walking posture and purely lateral impact can only be taken as a simplified setup for laboratory-based use. Whether it is the causation of the variation in results needs to be further investigated.

Human body model proved to be an efficient tool in identifying the contribution of pre-impact posture in kinematics and injury risk [11]. The computational approach was taken in this study for its capability of eliminating individual diversity of the samples in cadaveric tests. There have been many studies related to pedestrian safety protection with THUMS (Total Human Body Model for safety) [14][20][21]. In the present study, THUMS pedestrian model was modified to simulate full-scale vehicle-to-pedestrian impact with two pre-impact postures (standing and walking) under 3 velocities, including 25, 33 and 40 kph, to investigate the influence of velocity on pedestrian kinematics and the coupling interaction of velocity and pre-impact postures. Different velocities of vehicle were considered to investigate the potential influence in a wide range of the initial energy of the whole dynamic event. Since vehicle front-end structure is deformable and so its resistance and local deformation to the leg contact change with different velocities. Higher contact velocity results in flatter contour of the deformed front-end structure, and consequently the different contact surface for lower extremity affects the knee joint bending angle and other kinematical characteristic. Lower extremity kinematics of pedestrian was investigated, including knee joint spatial bending motion and pelvic spatial translation and rotation. The goal of this study is to analyze influence of pre-impact pedestrian posture on lower extremity kinematics and mechanisms, and possible effects to injury risks of lower extremity.


Model Setup

Two gait states were adopted in the current study, the original THUMS pedestrian model (Version 4.0.1) in a standing posture (Figure 1a) and the mid-stance walking posture modified from the standing model [9]. The posture adjustment was made by applying quasi-static loading at the extremities with the upper body fixed as rigid. All the long bones of the upper and lower extremities were rigidified, with the knee, elbow and other essential connecting joints left deformable. In the pre-simulations for generating different gait postures, the torso was also converted to rigid body to improve the computational efficiency. 1-D beam element was attached to the distal end of the target extremity. Free end of the beam element was given a prescribed slow motion to pull the extremity to the predetermined position. Unlike a prescribed motion is assigned to bone structure directly, such an indirect loading applying method can avoid interfere in joint motion. Final postures of the model are illustrated in Figure 1.

A detailed front-end model of a mid-size sedan, used in the experiments by Subit et al. (2008), was used in this study [12]. The front-end structures included engine hood, upper/lower grill, bumper assembly, lower stiffener and other structures relevant to pedestrian lower extremity impact. The A-pillar was rigidly fixed. The geometry of the sedan model and the global coordinate system definition were illustrated in Figure 2. The adopted model had already been validated by Watanabe et al. against the cadaveric test results including data of accelerometers and markers in their study [11].

In the present study, simulations were conducted using explicit LSDYNA software (LSTC, Livermore, CA, Version 971 R6.1.2). The whole simulation time for a single case was 100 ms and each took approximately 13 hours in physical time in a cluster of Intel Xeon E5-2650 processor with 64 cores.

The pedestrian was positioned 5 mm in front of the bumper central outline to mimic the instant prior to a typical vehicle-to-pedestrian collision. The pedestrian was in standing still posture or mid-stance gait with vehicle striking the left side. The ipsilateral (struck side) leg was moving forward in walking case. The ground was modeled by rigid wall and the friction coefficient was set to 0.2 between the pedestrian and the ground. Global gravity field was assigned to simulate the gravity effect. Braking effect and subsequent pitching motion of vehicle front-end were not taken into account.

Characterization of Pedestrian Kinematics

Displacements of several body regions were measured relative to the vehicle coordinate system to characterize the global kinematics, including the head center of gravity, the acromia, the spine (First thoracic vertebra T1 and eighth thoracic vertebra T8), top of the sacrum, great trochanter, femoral and tibial shaft, and tibial plateau (both the medial and lateral sides) (Figure 3).

In the post process, both ipsilateral and contralateral knee joint bending angles in the three directions were calculated by the relative rotation of the local coordinate systems defined on femur and tibia using the Grood-Suntay method [13]. The two local coordinate systems were defined based on nodal information in femoral condyle and tibial plateau areas, respectively, to represent the structure motion (Figure 4). This decision was made considering the much higher stiffness and less deformation of the femoral condyle and tibial plateau areas comparing to adjacent tissues. Relative bending angle between the two subjects in the three directions, flexion, abduction/adduction, tibial external/internal rotation, can be obtained by inverse-consine calculation of projection of femoral axis on the corresponding tibial axis, and the formula is referred to [13].

Similarly, pelvic spatial motion can be calculated with respect to the global coordinate system. Ligament stretching ratio was used as an indicator to monitor ligament injury risk. For each of the concerned ligament, the stretching ratio was calculated by summing up of piecewise length between adjacent nodes in linear consequence along its length.

Definition of Injury Predictors

In the present study, long bone fracture and ligament rupture were taken into consideration, since those were main injury patterns for lower extremities.

For knee ligament injury, ligament stretching ratio was selected as the injury predictor. Comparing to other stress/strain-based predictors, stretching ratio was of higher robustness, and the response would not be highly dependent on mesh quality, especially for those cases with large local deformation in knee ligaments. The ligament length was calculated by sum of piecewise element length in ligament fiber direction (Figure 5). Stretching ratio was the value of the elongation part divided by the original length.

For tibial fracture, plastic strain exceeding 1.5% may be regarded as the injury predictor and corresponding threshold in TFIUMS [19]. Von-Mises stress was not regarded as an injury predictor. However, it could reflect the local force load to the bone structure. Thus it was chosen to evaluate local contact force in tibia.

Element elimination of knee joint ligaments and bone structure was deactivated in the simulations.


Global Kinematics

The traces of the markers on the pedestrian body are shown in Appendix A. Table 1 lists the peak excursion values referring to vehicle coordinates in the X-Z plane of different body parts.

In high velocity case, generally the peak excursion was larger. The current cases lasted 100 ms, which did not cover the rebounding phase of the whole impact process for pedestrian. Thus, the peak value occurred at 100 ms.

Considering the influence of pre-impact postures on excursion, walking cases presented drastically larger excursion in right knee joint and heel in X direct. That was because there was no two leg interaction and the right leg could move towards vehicle. For pelvis and thorax, excursion in walking cases was slightly larger, about 5-10%.

Knee Joint Kinematics

The knee joint motion for standing and walking postures under vehicle velocity of 25 kph is illustrated in Figure 6 as a representative of the three impact severities. At 25 kph vehicle impact speed the curves exhibit more details as the lower speed corresponding with long impact process. On the contrary larger velocity case corresponds with shorter time in impact process, periods induced by different mechanisms cannot be obviously distinguished in the curves. All schematics of the knee joint kinematics curves are shown in Appendix B.

Knee joint presented large variation for different pedestrian preimpact postures.

For the ipsilateral knee joint, peak value of abduction from the two pre-impact postures were similar, i.e., 15.2 degree for standing (Figure 6a) and 15.6 degree for the walking posture (Figure 6b).

Load transfer pattern variation was revealed by difference in curve shapes for the two postures. The ipsilateral side in the standing posture exhibited a dual peak phenomenon in the abduction and flexion curve (Figure 6a). Mechanism for this is the contact of the two legs in knee joint area. Descent of the ipsilateral abduction curve was at the same instant when contralateral knee joint initiated to move around 25 ms. Flexion angle varied along with abduction in the curve shape regardless of pre-impact posture. This can be explained as flexion was a natural motion of knee joint and was easy to trigger during a complicated spatial motion. Therefore, it was initiated by the lower extremity to adjust itself into a more flexible and relaxed configuration. Excessive abduction was an abnormal motion from anatomical view due to external loading. When knee joint deviated from the initially perfect lateral loading condition to some kind of oblique bending condition, flexion would emerge to mitigate internal tensity in knee joint caused by lateral bending. Tibial external rotation exhibited a significant oscillation (Figure 6) and small peak value compared to flexion and abduction/adduction curves. Overall, the pre-impact posture did not present a significant influence on the ipsilateral knee joint motion. In the initial period the motion curves for the ipsilateral legs were similar before two legs started interaction. Difference in the external rotation curve was induced by the swing of ipsilateral leg in walking posture.

Two-leg interaction induced larger difference in kinematics of contralateral knee joint. For the ipsilateral knee joint, peak value of abduction from the two pre-impact postures were similar, i.e., 12.5 degree for standing (Figure 6a) and 20.1 degree for the walking posture (Figure 6b). There was a 40% difference. For pedestrian in standing posture, contact at the lateral side of the femoral condyle resulted in femoral internal and in-coronal-plane rotation. This led to an extension motion and the contralateral knee joint was loaded in a "locked" state (Figure 7). The extension was the consequence of local geometry around the contact location and such motion brought more constrained adduction motion of the contralateral knee joint, that is, the knee joint was loaded towards a more deflected direction from the normal state. For walking posture the contralateral leg impacts directly with the front-end structures without obstruction from the other leg. Prior to the contact, the femur head had already been loaded by acetabulum due to pelvic motion. Thus there was an original descent in the flexion and adduction curve. Peak value was also higher comparing to the ipsilateral side. As the thigh did not cling to vehicle surface, the far-side leg possessed larger space to deform. Thus the lateral bending angle in knee joint was larger for the walking case (Figure 8).

Pelvic Motion and Relation to Pre-Impact Postures

Pelvis played a predominant role in load transfer from lower extremities to upper torso [6]. In the present study, pre-impact posture of pedestrian induced differences in pelvic motion. As mentioned above, contralateral knee joint in standing case was "locked". Lower extremity motion was more constrained in the coronal plane. Walking would bring complicated nonplanar configuration to initial condition which can be viewed as a "disturbance" to the equilibrium of the lower extremity. Specifically, although the initial states of the pelvis were practically the same for the two postures, it exhibited significantly higher rotation along the body z-axis during the impact in walking posture. (Figure 9).

For pedestrian in standing posture, pelvis rotation about Y-direction and Z-direction were significant and the magnitudes were close in a large time range (Figure 9a, c, e). Walking posture presented a significant higher rotation about the Z-direction rather than X-direction and Y-direction. The reverse Z-direction rotation of pelvis would turn the human body to a position with larger contact area with the vehicle front-end. In 40 kph cases, the largest Z-direction rotation was 21.6 degree for standing and 29.8 degree for walking. There was a 27.5% difference. Hip joint got closer to vehicle. Pelvis was consequently more backward in the impact direction, indicating that it acquired less momentum from the vehicle (Figure 10). Less amount of momentum transfer corresponded with smaller contact force which was beneficial to pedestrian safety. The pedestrian pelvis displacement in the global X-direction is shown in Figure 11. The marked point was at the conjunction point between sacrum and L5 vertebra. Thus the X displacement of the specific point partially reflected the inertial effect of the upper torso, that is, how much the upper torso got involved in the X-direction motion. Take 40 kph cases for example, the X-direction displacement was -368 mm for standing case and -334 mm for walking case. There is a 10.2% difference of X-direction displacement, which means that locally the lumbar components acquired more momentum in the standing case.

The results were consistent with what was revealed by the rotation curves. The relatively larger Z direction rotation and smaller X displacement in the walking posture implies less constraint on the pelvis motion and a potentially lower injury risk.

Injury of Lower Extremities

Abduction/adduction of knee joint (conventionally referred to as the lateral bending angle) was chosen to model the relative movement between femur and tibia besides the knee joint kinematics itself in the present study. They were also considered as a main injury predictor, especially for the collateral ligaments [15][16][17]. This study was intended to focus on the lower extremity kinematics, therefore specific knee joint kinematical and injury parameters such as lateral shearing displacement, tibial anterior/posterior displacement, ligament maximum principal strain and local stress of the other connection soft tissues were not included. Excessive abduction/adduction tended to cause large collateral ligament stretching in the tensile side or and may result in ligament rupture. Table 2 listed the maximum ligament stretching ratio and occurrence time for ipsilateral MCL and contralateral LCL. Generally, walking posture presented higher stretching ratio corresponding to higher risk of rupture reflected by maximum stretching ratio of ligaments in Table 2. For example, in 40 kph cases, ipsilateral MCL maximum stretching ratio was 18.8% for walking case and 16.6% for standing case. Overall collateral ligament stretching in the walking cases was 20-30% larger than that in the standing cases. The stretching ratio was approximately linearly proportional to the lateral bending angle due to the anatomical features of knee structure. Shearing displacement and axial tension of knee joint also contributed to ligament stretching. As the ligaments served as main connection tissue in knee joint structure, their mechanical behavior was dependent on knee joint local kinematics, which was highly relevant to pre-impact postures. Especially for the contralateral LCL the stretching ratio exceeded 30% under 40 kph velocity. Generally regardless of pre-impact posture, LCL stretching ratio increases with velocity.

The phenomenon that contralateral LCL presented larger stretching ratio and corresponding higher injury risk relative to ipsilateral MCL was already addressed by Kerrigan in his cadaveric tests [10], LCL had a smaller cross section area and was weaker in structure. Thus the stretching ratio was larger under similar axial force along the fiber direction. And the contralateral leg undertook the axial tension induced by pelvis rotation. The pulling motion in vertical direction added on the axial force in knee joint ligaments. Together with lateral bending motion, the axial tension force in contralateral LCL was magnified. Such phenomena were not observed in previous cadaveric tests and field data, probably because the boundary condition was not perfect lateral impact for the realistic accident and there might be rotation during the impact process. The mechanism behind the difference of response in ipsilateral and contralateral legs is still under investigation.

Component test in current regulations towards pedestrian lower extremity protection was established to represent the impact of ipsilateral leg. For the contralateral leg impact, axial tension and interaction of the two legs should be taken into consideration in future.

Another injury pattern always emerged in accidents was tibia fracture, which is commonly caused by instantaneous impact load from bumper and the subsequent bending load during continuous vehicle intrusion. Injury risk of tibia could be assessed by local stress in the shaft area. The Von-Mises stress history of the element with the global maximum stress value in the ipsilateral tibia was shown in Figure 12.

On the contrary, standing posture always presented larger tibial stress, especially in low velocity case. In 25 kph cases, the peak stress in standing posture (75 MPa) was 25% larger than that in walking case (60 MPa). The phenomenon could be attributed to the two-leg interaction. The curves showed a dual-peak shape in which the first one was due to bumper impact and the second one was induced by tibial lateral bending. With overlap of two legs in standing posture, the contralateral leg inertia prevented the ipsilateral tibia from freely doing the X direction motion and thus more mass was involved in the system, resulting in higher internal force in lower extremities

The larger deformation brought by more unconstrained motion in walking posture generally led to higher injury risk in knee joint ligaments but it can decrease the stress in hard tissues, like bones. For pedestrian safety protection design, the conflict of different injury risk distribution induced by pre-impact postures should receive more attention.


Model Set Up

The present study mainly focused on the influence of pedestrian initial posture on lower extremity kinematics. The boundary condition was to reflect the vehicle-to-pedestrian impact rather than reconstructing cadaveric tests or realistic accident scenario. Thus, boundary condition of the impact simulation was set up based on generic impact scenario. Braking effect and subsequent pitching motion were not considered in the current study to avoid interfere from over complicated input parameters.

Limitation of Current Study

In the present study, a 50 percentile pedestrian model was used to investigate the influence of pedestrian pre-impact posture on lower extremity kinematics. Stature of pedestrian is commonly believed to significantly affect the contact location for pelvis and knee joint [6]. Difference in contact location will subsequently affects local stress distribution and the kinematics. Parametric human body modeling is promising to replicate the geometry over a certain range of population by adapting the model into different body size. Further study about the influence of pre-impact postures with a range of anthropometries needs to be conducted.

The current study took standing and mid-stance as the two representative postures to investigate the influence of pre-impact postures. For more gait postures that can be regarded as transition between standing and mid-stance gait, the present study had not taken those into the simulation matrix, some of the conclusions from this study may shed some lights on further studies on more gaits in the future.

Differences in response brought by changes in postures can be characterized in several aspects. The first was that the active mass getting involved in the impact process. For the mid-stance gait, only small overlap between two legs occurred and the overlap should be getting larger for those transition gaits. In the standing posture, the two legs completely overlapped with each other. The current results revealed that smaller active mass in lower extremities for mid-stance gait would reduce peak value in tibia stress, which is applicable to the transition gaits relative to standing posture. The other aspect of difference induced by changes in gaits was that the motion of pelvis. The pelvis rotation about the Z-axis was released in the walking posture. The spatial rotation was observed which was analyzed in the results section. For those transition gaits the pelvis motion should be similar with that in the mid-stance gait since the two legs would not present perfectly planar motion as that for the standing posture.

Knee joint motion was hard to be simply generalized from the standing and the mid-stance to other gaits because of the complicated analytical features in knee joint. Detailed influence of more gaits should be investigated in future studies.

Injury threshold of ligament stretching ratio was in a large range in current test results [18]. Failure thresholds present obvious individual diversity, and motion after failure in local tissue cannot be validated. Thus failures and failure thresholds were not defined in the current model. Without use of element deletion, tissue failure could not be reflected during the simulation. Failure behavior of tissues is not a focus of the current. Response before the failure instant (i.e., before the predictor exceeded the threshold) was realistic. After the instant, the failure should occur and the model was then not able to predict a realistic motion of lower extremities. Under many circumstances in this study, lower extremity was not in dangerous situation. Peak value for each predictor varying with different impact scenario could relatively reflect injury trend and provided researchers with insight into the influence of different pre-impact postures. For future studies, interaction of failure and subsequent kinematics with different preimpact postures needs to be investigated.

The current model did not include muscle activation. Muscle activation can affect the stretch of knee ligaments during impact since the muscle in the knee joint area will burden part of the load from femur to tibia. Different impact postures induces variation in muscle activity which should be taken into consideration in future study.


This study investigated the influence of pre-impact postures on the kinematics of pedestrian lower extremities in vehicle-to-pedestrian impact. A 50th adult pedestrian model in both standing and walking postures were taken into account under different impact velocities. Global kinematics of pedestrian was studied. Rotation of the knee joint around three axis was calculated and pelvic translational and rotational motion were analyzed. For the walking case, the contralateral knee joint lateral bending angle was 40% larger than the standing case. Stretching ratio for all ligaments was 20-30% larger, and pelvic Z-direction rotation was 27.5% larger in the walking cases. On the contrary, pelvic X-direction translation was 10.2% smaller and tibial stress was 25% smaller in the walking case. Pedestrian in walking posture exhibited larger knee bending angle and pelvic rotation due to isolated single-leg contact with vehicle and nonplanar characteristic from the leg swing. The walking posture increased the injury risk of soft connection tissue about 20-30% and reduced the internal force in bone structure about 25% regardless of impact severity. The trade-off of injury risks induced by kinematics with different pre-impact postures was a challenge for vehicle front-end structure design. Further research efforts are necessary to include more loading scenarios and find a balance between protection of long bones and soft tissues.


[1.] Chidester. A., Isenberg, R., "Final report: The pedestrian crash data study. Society of Automotive Engineers", 2001.

[2.] Untaroiu C., Kerrigan, J., Kam, C., et al., "Correlation of strain and loads measured in the long bones with observed kinematics of the lower limb during vehicle-pedestrian impacts," Stapp car crash journal 51: 433-466, 2007.

[3.] Ishikawa, H., Kajzer, J., and Schroeder, J., "Computer simulation of impact response of the human body in car-pedestrian accidents," SAE Technical Paper 933129, 1993, doi: 10.4271/933129.

[4.] Arnoux, P., Cavallero, C., Chabrand, P., Brunet, C., "Knee ligament failure under dynamic loadings," International Journal of Crashworthiness, 7:3, 255-268, 2002, doi:10.1533/cras.2002.0218.

[5.] Kerrigan, J., Bhalla, K., Madeley, N., Funk, J. et al., "Experiments for Establishing Pedestrian-Impact Lower Limb Injury Criteria," SAE Technical Paper 2003-01-0895, 2003, doi: 10.4271/2003-01-0895.

[6.] Kerrigan, J., Crandall, J., Deng, B., "Pedestrian kinematic response to mid-sized vehicle impact," International journal of vehicle safety, 2(3): 221-240, 2007.

[7.] Kerrigan, J., Crandall, J., Deng, B., "A comparative analysis of the pedestrian injury risk predicted by mechanical impactors and post mortem human surrogates," Stapp car crash journal, 52: 527-567, 2008.

[8.] Anderson, R., McLean, A., Dokko, A., "Determining accurate contact definitions in multi-body simulations for DOE-type reconstruction of head impacts in pedestrian accidents," Proceedings of the 19th International Technical Conference on the Enhanced Safety of Vehicles. 6-9, 2005.

[9.] Untaroiu, C., Meissner, M., Crandall, J., et al., "Crash reconstruction of pedestrian accidents using optimization techniques," International Journal of Impact Engineering, 36(2): 210-219, 2009, doi: 10.1016/j.ijimpeng.2008.01.012.

[10.] Kerrigan, J., Arregui-Dalmases, C., Foster, J., et al., "Pedestrian injury analysis: field data vs. laboratory experiments," Proceedings of the International Research Council on the Biomechanics of Injury conference, 40: 672-689, 2012.

[11.] Watanabe, R., Katsuhara, T., Miyazaki, H., Kitagawa, Y., Yasuki, T. "Research of the Relationship of Pedestrian Injury to Collision Speed, Car-type, Impact Location and Pedestrian Sizes Using Human FE Model (THUMS Version 4)," Stapp Car Crash Journal, 56: 269-321, 2012.

[12.] Subit, D., Kerrigan, J., Crandall, J., et al., "Pedestrian-vehicle interaction: kinematics and injury analysis of four full scale tests," Proceedings of the International Research Council on the Biomechanics of Injury conference, 36: 275-294, 2008.

[13.] Grood, E., Suntay, W., "Ajoint coordinate system for the clinical description of three-dimensional motions: application to the knee," Journal of biomechanical engineering, 105(2): 136-144, 1983. doi: 10.1115/1.3138397.

[14.] Han, Y., Yang, J., Mizuno, K., Matsui, Y., "Effects of vehicle impact velocity, vehicle front-end shapes on pedestrian injury risk," Traffic injury prevention, 13(5): 507-518, 2012, doi: 10.1080/15389588.2012.661111

[15.] Kajzer, J., Schroeder, G., Ishikawa, H., Matsui, Y. et al., "Shearing and Bending Effects at the Knee Joint at High Speed Lateral Loading," SAE Technical Paper 973326, 1997, doi: 10.4271/973326.

[16.] Kajzer, J., Matsui, Y., Ishikawa, H., Schroeder, G. et al., "Shearing and Bending Effects at the Knee Joint at Low Speed Lateral Loading," SAE Technical Paper 1999-01-0712, 1999, doi: 10.4271/1999-01-0712.

[17.] Mo, F., Masson, C., Cesari, D., Arnoux, P., "Coupling Lateral Bending and Shearing Mechanisms to Define Knee Injury Criteria for Pedestrian Safety", Traffic Injury Prevention, 14(4): 378-386, 2013, DOI: 10.1080/15389588.2012.721146.

[18.] Mo, F., Arnoux, P., Cesari, D., Masson, C., "The failure modelling of knee ligaments in the finite element model", International Journal of Crashworthiness, 17(6): 630-636, 2012, doi: 10.1080/13588265.2012.704194.

[19.] Yasuki, T., Yamamae, Y., "Validation of kinematics and lower extremity injuries estimated by total human model for safety in SUV to pedestrian impact test," Journal of Biomechanical Science and Engineering, 5(4). 340-356, 2010, doi: 10.1299/ibse.5.340.

[20.] Chen, H., Poulard, D., Crandall, J., Panzer, M., "Pedestrian Response with Different Initial Positions During Simulated Impact with a Mid-Sized Sedan," In 24th International Technical Conference on the Enhanced Safety of Vehicles (ESV) (No. 15-0391), 2015

[21.] Han, Y., Yang, J., Nishimoto, K., Mizuno, K., Matsui, Y., Nakane, D., Hitosugi, M., "Finite element analysis of kinematic behaviour and injuries to pedestrians in vehicle collisions," International Journal of Crashworthiness, 17(2): 141-152, 2012, doi: 10.1080/13588265.2011.632243.


Jisi Tang is now a Ph.D candidate in Department of Automotive

Engineering, Tsinghua University

Address: Automobile Crash Lab, Tsinghua University, Beijing,

China, 100084


This research is sponsored by Toyota Motor Corporation. The authors would like to thank LSTC for providing us the educational license of LS-DYNA. We also appreciate that Arup provides us the academic license of OASYS software.


Appendix A

Appendix B

Jisi Tang, Qing Zhou, and Bingbing Nie Tsinghua University

Tsuyoshi Yasuki and Yuichi Kitagawa Toyota Motor Corporation

Table 1. Peak excursions of different body parts

            Peak excursion (mm)
Body parts  25 kph                        33 kph
            Standing             Walking  Standing  Walking

Head X      570                  572       719       729
Head Z      -92.1                -65.1    -127      -120
T8 X        527                  540       661       646
T8 Z        -84.3                -60.1    -118      -104
Pelvis X    330                  371       407       421
Pelvis Z    -23.1                -22.5     -21.5     -23.5
Knee R X     82.5                210       109       187
Knee R Z    151                   77.5     220       146
Heel R X    -15.4                265       -43.5     204
Heel R Z    152                   96.5     241       140

Body parts  40 kph
            Standing  Walking

Head X       850       845
Head Z      -169      -151
T8 X         736       741
T8 Z        -143      -140
Pelvis X     470       473
Pelvis Z      17.6      19.2
Knee R X     131       233
Knee R Z     289       201
Heel R X     -94.7      86.3
Heel R Z     323       201

Table 2. Maximum stretching ratio of collateral ligaments for all cases

                            Ipsilateral MCL
Velocity (kph) and posture  Max stretching ratio (%)  Occurrence time

25 standing                 10.9                      56
25 walking                  13.3                      42
33 standing                 13.4                      46
33 walking                  16.6                      38
40 standing                 16.6                      42
40 walking                  18.8                      34

                            Contralateral LCL
Velocity (kph) and posture  Max stretching ratio  Occurrence time (ms)

25 standing                 14.9                  76
25 walking                  22.3                  78
33 standing                 19.8                  56
33 walking                  26.3                  68
40 standing                 24.7                  50
40 walking                  31.2                  60
COPYRIGHT 2016 SAE International
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2016 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Tang, Jisi; Zhou, Qing; Nie, Bingbing; Yasuki, Tsuyoshi; Kitagawa, Yuichi
Publication:SAE International Journal of Transportation Safety
Article Type:Report
Date:Jul 1, 2016
Previous Article:A semi-automated approach to real world motor vehicle crash reconstruction using a generic simplified vehicle buck model.
Next Article:The wagging foot of uncertainty: Data collection and reduction methods for examining foot pedal behavior in naturalistic driving.

Terms of use | Privacy policy | Copyright © 2021 Farlex, Inc. | Feedback | For webmasters