Investigation of a Relationship Between External Force to Shoulder and Chest Injury of WorldSID and THUMS in 32 km/h Oblique Pole Side Impact

Transcription

1 Traffic Injury Prevention ISSN: (Print) X (Online) Journal homepage: Investigation of a Relationship Between External Force to Shoulder and Chest Injury of WorldSID and THUMS in 32 km/h Oblique Pole Side Impact Shinobu Tanaka, Shigeki Hayashi, Satoshi Fukushima & Tsuyoshi Yasuki To cite this article: Shinobu Tanaka, Shigeki Hayashi, Satoshi Fukushima & Tsuyoshi Yasuki (2013) Investigation of a Relationship Between External Force to Shoulder and Chest Injury of WorldSID and THUMS in 32 km/h Oblique Pole Side Impact, Traffic Injury Prevention, 14:sup1, S64-S76, DOI: / To link to this article: Toyota Motor Corporation View supplementary material Published online: 01 Aug Submit your article to this journal Article views: 570 View related articles Full Terms & Conditions of access and use can be found at Download by: [ ] Date: 26 December 2016, At: 05:59

2 Traffic Injury Prevention (2013) 14, S64 S76 Published with license by Taylor & Francis ISSN: print / X online DOI: / Investigation of a Relationship Between External Force to Shoulder and Chest Injury of WorldSID and THUMS in 32 km/h Oblique Pole Side Impact SHINOBU TANAKA, SHIGEKI HAYASHI, SATOSHI FUKUSHIMA, and TSUYOSHI YASUKI Toyota Motor Corporation, Toyota, Aichi, Japan Received 15 March 2013, Accepted 12 April 2013 Objective: This article describes the chest injury risk reduction effect of shoulder restraints using finite element (FE) models of the worldwide harmonized side impact dummy (WorldSID) and Total Human Model for Safety (THUMS) in an FE model 32 km/h oblique pole side impact. Methods: This research used an FE model of a mid-sized vehicle equipped with various combinations of curtain shield air bags, torso air bags, and shoulder restraint air bags. As occupant models, AM50 WorldSID and THUMS AM50 Version 4 were used for comparison. Results: The research investigated the effect of shoulder restraint air bag on chest injury by comparing cases with and without a shoulder side air bag. The maximum external force to the chest was reduced by shoulder restraint air bag in both WorldSID and THUMS, reducing chest injury risk as measured by the amount of rib deflection, number of the rib fractures, and rib deflection ratio. However, it was also determined that the external force to shoulder should be limited to the chest injury threshold because the external shoulder force transmits to the chest via the arm in the case of WorldSID and via the scapula in the case of THUMS. Because these results show the shoulder restraint air bag effect on chest injury risk, the vent hole size of the shoulder restraint air bag was changed for varying reaction forces to investigate the relationship between the external force to the shoulder and the risk of chest injury. In the case of THUMS, an external shoulder force of 1.8 kn and more force from the shoulder restraint air bag was necessary to help prevent rib fracture. Increasing external force applied to shoulder up to 6.2 kn (the maximum force used in this study) did not induce any rib or clavicle fractures in the THUMS. When the shoulder restraint air bag generated external force to the shoulder from 1.8 to 6.2 kn in THUMS, which were applied to the WorldSID, the shoulder deflection ranged from 35 to 68 mm, and the shoulder force ranged from 1.8 to 2.3 kn. Conclusions: In the test configuration used, a shoulder restraint using the air bag helps reduce chest injury risk by lowering the maximum magnitude of external force to the shoulder and chest. To help reduce rib fracture risk in the THUMS, the shoulder restraint air bag was expected to generate a force of 3.7 kn with a minimum rib deflection ratio. This corresponds to a shoulder rib deflection of 60 mm and a shoulder load of 2.2 kn in WorldSID. Supplemental materials are available for this article. Go to the publisher s online edition of Traffic Injury Prevention to view the supplemental file. Keywords: biomechanics, crash dummies, chest injury, side impact, side air bags, crash worthiness Introduction Many countries around the world have adopted regulations or assessments for side impacts that use anthropomorphic dummies to evaluate vehicle occupant protection performance. Development of the 50th percentile American male worldwide harmonized side impact dummy (WorldSID) began in 1997 under the auspices of the governments and specialist institu- Toyota Motor Corporation Address correspondence to Shinobu Tanaka, Toyota Motor Corporation, 1, Toyota-cho, Toyota, Aichi , Japan. shinobu@tanaka.tec.toyota.co.jp tions that make up the WorldSID Task Group as a global standard test dummy designed to accurately simulate the human body. An evaluation performed in accordance with ISO/TR 9790, comprising impact tests on body parts such as the head, chest, pelvis, and so on, confirmed that WorldSID has a good biofidelity score of 8/10 (Sherer et al. 2009). Starting with the 2015 European New Car Assessment Programme (Euro NCAP), WorldSID is expected to be adopted in a growing number of regulations and assessments around the world. In addition, the injury criteria for evaluations using WorldSID are currently being discussed by Euro NCAP and the Global Road Safety Partnership. Haenchen et al. (2004) reported that 49.1 percent of collision accidents in Germany since 1991 that resulted in a serious

3 Oblique Side Impacts 65 Fig. 1. Shoulder pendulum test of WorldSID (color figure available Fig. 3. Abdomen pendulum test of WorldSID (color figure available injury (i.e., an Abbreviated Injury Scale score of 3 or more) were due to a collision with a tree, telephone pole, and so on (these types of objects are referred to in this article as poles). Furthermore, Otte et al. (2008) reported that, in collision accidents in Germany between 1999 and 2006, the probability of injury from a pole side impact was approximately 40 percent, a relatively high figure compared to other types of accidents. It was also reported that 32 percent of these injuries were to the chest, which was the second highest percentage after injuries to the head. This shows that the reduction in chest injuries in pole side impacts is an important issue for lowering the number of people injured or killed in actual side impact accidents. Hayashi et al. (2008) used the Total Human Model for Safety (THUMS) finite element (FE) model to demonstrate the effectiveness of side air bags in reducing rib fractures and strain on internal organs in side impacts. In addition, Smith et al. (2011) used THUMS to verify the effectiveness of shoulder restraint using a seat shoulder support to reduce the risk of chest injury in stock car racing. Although restraining the shoulder using the seat or air bags is regarded as a potential way of reducing chest injury risk in pole side impacts, there is discussion as to whether they might increase shoulder or chest injury by applying force to the shoulder. For this research, 32 km/h oblique pole side impacts were simulated using a WorldSID FE model and a human FE model in vehicles equipped with shoulder and torso side air bags in order to investigate the relationship between chest injury risk and external force application to the shoulder. Methods Model Validation DYNAmore WorldSID Ver Figures 1 to 4 compare the results of the WorldSID FE model with the results of a validation test using a WorldSID owned by Toyota Motor Corporation using a pendulum. Of the verification tests described by the user manual (DYNAmore GmbH 2010), the figures show examples of results for the shoulder, chest (with arm), abdomen, and pelvis. The seating posture of the dummy as well as the impact position and speed of the pendulum on the dummy were in accordance with the manual. The pink lines show the test results, and the blue lines show the FE model results. The waveform and peak impactor reaction force at the shoulder were generally consistent between the computer aided engineering (CAE) model and the test results. Although the peak deflection of the middle thorax rib in the CAE model was slightly higher than in the test, both were within the validation corridor. As a result, the difference between these 2 results can be regarded as insignificant. The peak deflection in the CAE model was also lower than the test results for the lower abdomen rib. However, because this research is examining chest injury, this difference was judged to be insignificant. The peak impactor reaction force at the pelvis was generally consistent with the test results. Fig. 2. Chest pendulum test of WorldSID (color figure available Fig. 4. Pelvis pendulum test of WorldSID (color figure available

4 66 Tanaka et al. Fig. 5. Lateral shoulder impact test by Sabine et al. (2004) (color figure available THUMS THUMS is a human body FE model. The THUMS simulates a 50th percentile American male with a height of 175 cm and a weight of 77 kg. THUMS Version 4 used in this study included not only the skeleton but also the internal organs such as lung, heart, and so on (Toyota Motor Corporation 2011). Watanabe et al. (2011) reported that the dynamic chest response to the lateral impact input corresponds to the corridor of postmortem human subject (PMHS) experiments performed by Viano (1989). In addition, this model can predict rib fracture by eliminating elements when strain levels exceed a specified value. The number of rib fractures during the chest impact was also verified by Watanabe et al. (2011). In addition to the validation of the chest, shoulder validation is necessary in this study. Figure 5 shows the model simulating a lateral shoulder impactor test reported by Sabine at al. (2004) using a PMHS. In the test, a 24.8-kg flat rigid impactor ( mm) was impacted into the lateral side of the shoulder at an initial velocity of 3.0 to 4.0 m/s and the impact force was measured. The tests were performed on a total of 4 PMHS to create a force response corridor at each velocity. A corridor of the impact at 4.5 m/s was obtained by scaling the original corridors. The force response of THUMS was plotted to the 4.5 m/s velocity corridor in Figure 5. The impactor force in the simulation was the contact force between the impactor and the shoulder in the THUMS. The horizontal axis shows time and the vertical axis shows the impactor force (Figure 5). Simulation results fit in the corridor except the first peak at 7 ms. It was estimated that the first peak was generated mainly from inertia force of the THUMS shoulder. Overall, the simulation model appears to be capable of predicting the lateral impact response of the human body. Fig km/h oblique pole side impact FE model (color figure available Vehicle Model The base FE models were validated by comparison with the results of a 32 km/h oblique pole side impact test using the validated AM50 WorldSID seated in a mid-sized vehicle equipped with curtain shield and torso air bags. The FE modeling was carried out following the methods reported by Kojima et al. (2004) for the vehicle body and Fukushima et al. (2008) for the air bags and other restraint devices. Figure 6 and Table 1 show an outline of the test conditions and FE models, respectively. The FE analysis was carried out using LS-DYNA (Ver. 971, Livermore Technology Corp., Livermore, CA). Figures 7 to 9 compare the test and FE simulation results for the residual deformation of the front door contacted by the dummy. The cross section A-A graph shows a front view of the dummy torso side section. Cross section graphs B-B and C-C show top views of the shoulder and hip point height sections, respectively. The pink lines show the test results, and the blue lines show the FE model results. The simulated results using the FE model were generally consistent with the test results in terms of the shape and amount of deformation for each section. Figure 10 compares the time history of the injury criteria for each part of WorldSID as obtained in the test and Table 1. Test conditions of this study Collision speed Collision angle Vehicle type Dummy Dummy seating Torso air bag Curtain shield air bag 32 km/h 75 Passenger vehicle WorldSID AM50 Ver.5.4 Prototype A Prototype α Fig. 7. Comparison of body deformation (cross section A-A) (color figure available

5 Oblique Side Impacts 67 Fig. 8. Comparison of body deformation (cross section B-B) (color figure available calculated using the FE model. The pink lines show the test results, and the blue lines show the FE model results. The time history of the shoulder force, middle thorax rib deflection, and pubic force are generally consistent in terms of both waveform and maximum value. However, slight differences were observed in the time history of the abdomen lower rib deflection restrained by the torso air bag and the door trim. The deflection calculated by the FE model was lower than that observed in the test because breakage and other behavior of plastic door trim pieces was not adequately simulated, which caused a lower deformation than that observed in the test. However, the dummy measurements were accurately simulated by the FE model. Pole Side Impact Simulations This research used an FE model of a 32 km/h oblique side impact validated with test results to investigate the effect of shoulder restraint by air bags on WorldSID and THUMS chest injury. Air Bag Specifications General torso air bags cover the region from shoulder to abdomen. But in this study, 2 air bags were created to clearly isolate the effect of shoulder restraint by air bags on chest injury: one that mainly restrains the chest and abdomen (i.e., a torso air bag) and one that mainly restrains the shoulder (i.e., a shoulder restraint air bag). Figure 11 shows these air bags and Table 2 lists the investigation combinations. Occupant Seating Posture and Position The front seat placement and WorldSID seating state was set in accordance with Seating Procedure Ver. 5.4 issued by the In- Fig. 10. Comparison of injury response (color figure available Fig. 9. Comparison of body deformation (cross section C-C) (color figure available Fig. 11. Equipped shoulder restraint air bag model (color figure available

6 68 Tanaka et al. Table 2. Combination of torso air bag and shoulder restraint air bag Case Dummy Torso Shoulder 1 WorldSID B None 2 B C 3 THUMS B None 4 B C Fig. 13. Contact region of the shoulder (color figure available Fig. 12. Sitting posture and position of WorldSID and THUMS Ver. 4. Shoulder Injury Risk The shoulder injury risk was evaluated for THUMS with and without clavicle fracture and for WorldSID using shoulder force measurement. The clavicle fracture criteria for THUMS was set at a fracture strain value of 3.0 percent for a 30- to 40-year-old as suggested by Watanabe et al. (2012). Petitjean et al. reported that shoulder force is a index of shoulder injury risk for WorldSID. ternational Organisation for Standardisation (2011). THUMS was set with the same hip point as WorldSID (Table 3) without changing the placement of the driver s seat. THUMS was given the standard seating posture proposed by the University of Michigan Transportation Research Institute (UMTRI 2013). Figure 12 shows the seating posture of WorldSID and THUMS. Chest Injury Risk For WorldSID, the chest injury risk was evaluated using the deflection of the upper, middle, and lower thorax ribs. For THUMS, the evaluation used number of rib fractures and the area around the left lung with a principal strain of 20 percent or more. Definition of External Force Applied to the Shoulder and Chest The external force applied to the shoulder is defined as the force of the contact between the dummy and the air bags and door trim in the regions shown in Figure 13. For WorldSID, this is the region above the lower edge of the shoulder rib in the side view when the shoulder and arm are set in the normal position. For THUMS, the shoulder is defined as the region above the armpit when the arm is set flat to the torso. The external force applied to the chest is defined as the force of contact between the dummy and the air bags and door trim in the regions shown in red in Figure 14. For WorldSID, this is the contact force between the upper and lower thorax ribs. For THUMS, the region above the lowest rib in an upright posture (i.e., the 10th rib) is defined as the chest and the shoulder is omitted from the chest region. Results Comparison of Restraint Timing at Each Occupant Region by Air Bags and Other Occupant Protection Devices WorldSID Figure 15 shows the restraint start timings by a torso air bag, door trim, and shoulder restraint air bag on each occupant region. It also compares the bottoming-out timing of the torso air bag with and without the shoulder restraint air bag. The Table 3. Specification of sitting posture and hip point Dummy Seating Hip point WorldSID Seating procedure Ver. 5.4 In accordance with Ver. 5.4 THUMS UMTRI Same as WorldSID Fig. 14. Contact region of the chest (color figure available

7 Oblique Side Impacts 69 Fig. 15. Timing of starting restraint of WorldSID (color figure available Fig. 16. Timing of starting restraint of THUMS (color figure available images in the figure show front views of the A-A cross section, which passes through the longitudinal center of the World- SID ribs. The blue lines show the WorldSID body, the red lines show the torso and shoulder restraint air bags, and the green lines show the surface of the door trim. Without the shoulder restraint air bag, the chest and abdomen were restrained first by the torso air bag 15 ms after the impact. Then the shoulder and pelvis were restrained by the door trim from 25 and 30 ms, respectively. In this case, the torso air bag bottomed out at approximately 35 ms. In contrast, with the shoulder restraint air bag, although there was no change in the restraint start timing for the chest, abdomen, and pelvis, the shoulder was restrained earliest at 12.5 ms. In this case, the torso air bag bottomed out at approximately 40 ms. Therefore, when the shoulder restraint air bag was present, the torso air bag restrained the chest for approximately 5.0 ms longer than without the shoulder restraint air bag. THUMS Figure 16 shows the restraint start timings by the air bags and door trim on each occupant region. It also compares the bottoming-out timing of the torso air bag with and without a shoulder restraint air bag. Without the shoulder restraint air bag, the chest and abdomen were restrained first by the torso air bag 17.5 ms after the impact. Then the shoulder and pelvis were restrained by the door trim from 27.5 and 35 ms, respectively. In this case, the torso air bag bottomed out at approximately 37.5 ms. In contrast, with the shoulder restraint air bag, although there was no change in the restraint start timing for the chest, abdomen, and pelvis, the shoulder

8 70 Tanaka et al. Fig. 17. External force to shoulder of WorldSID (color figure available was restrained earliest at 12.5 ms. In this case, the torso air bag bottomed out at approximately 42.5 ms. Therefore, when the shoulder restraint air bag was present, the torso air bag restrained the chest for approximately 5.0 ms longer than without the shoulder restraint air bag. Comparison of External Force Applied to Shoulder and Chest WorldSID Figure 17 compares the external force as described Figures 13 and 14 applied to the shoulder with and without a shoulder restraint air bag. The pink line shows the force without the shoulder restraint air bag and the blue line shows the force with the shoulder restraint air bag. Without the shoulder restraint air bag, the external force applied to the shoulder increased dramatically at approximately 40 ms, reaching a maximum value of approximately 11.5 kn. At 40 ms, the door trim bottomed out after applying a restraining force to the shoulder from 25 ms. In contrast, when the shoulder restraint air bag was present, a restraining force was applied by the air bag to the shoulder from 12.5 ms. The force also increased at approximately 40 ms when the door trim bottomed out. However, the high force application generated without the shoulder restraint air bag after the door trim bottomed out did not occur. Therefore, with the shoulder restraint air bag, a maximum external force of approximately 3.2 kn was generated at about 18.0 ms, which was roughly 30 percent of the maximum force without the shoulder restraint air bag. Figure 18 compares the external force applied to the chest with and without a shoulder restraint air bag. The pink line shows the force without the shoulder restraint air bag and the blue line shows the force with the shoulder restraint air bag. Without the shoulder restraint air bag, the external force applied to the chest increased at approximately 38 ms, reaching a maximum value of approximately 3.3 kn. At 33 ms, the door trim bottomed out after the torso air bag started to apply a restraining force to the chest from 12 ms. In contrast, when the shoulder restraint air bag was present, a restraining force was applied by the torso air bag to the chest from 12 ms. The force also increased at approximately 34 ms when the door trim bottomed out, reaching a maximum value of approximately Fig. 18. External force to chest (color figure available 3.0 kn. Therefore, the maximum external force applied to the chest when the shoulder restraint air bag was present was about 90 percent of the force without the shoulder restraint air bag. THUMS Figure 19 compares the external force applied to the shoulder with and without a shoulder restraint air bag. The pink line shows the force without the shoulder restraint air bag and the blue line shows the force with the shoulder restraint air bag. Without the shoulder restraint air bag, the external force applied to the shoulder continued to increase until the kinetic energy of the occupant was completely absorbed, reaching a maximum value of approximately 4.6 kn. This continuous increase began after the door trim applied a restraining force to the shoulder from 22.5 ms. In contrast, when the shoulder restraint air bag was present, a restraining force was applied by the air bag to the shoulder from 12.5 ms. Subsequently, the force increased at approximately 40 ms when the door trim bottomed out, reaching a maximum value of approximately 3.2 kn. Therefore, with the shoulder restraint air bag, the maximum external force applied to the shoulder was roughly 70 percent of that without the shoulder restraint air bag. Figure 20 compares the external force applied to the chest with and without a shoulder restraint air bag. The pink line shows the force without the shoulder restraint air bag and the blue line shows the force with the shoulder restraint air bag. Fig. 19. External force to shoulder (color figure available

9 Oblique Side Impacts 71 Fig. 20. External force to chest (color figure available Without the shoulder restraint air bag, the external force applied to the chest increased at approximately 34.0 ms, reaching a maximum value of approximately 4.2 kn. At 34.0 ms, the door trim bottomed out after the torso air bag started to apply a restraining force to the chest from 17.5 ms. In contrast, when the shoulder restraint air bag was present, a restraining force was applied by the torso air bag to the chest from 17.5 ms. The force also increased at approximately 35 ms when the door trim bottomed out, reaching a maximum value of approximately 2.7 kn. Therefore, the maximum external force applied to the chest when the shoulder restraint air bag was present was roughly 64 percent of the force without the shoulder restraint air bag. Comparison of Spinal Displacement WorldSID Figure 21 compares the lateral displacement of the WorldSID spine (T1, T4, and T12) with and without a shoulder restraint air bag. The comparison was made at 44 ms, at which the maximum spinal displacement was reached in the case without the shoulder restraint air bag. Adopting the shoulder restraint air bag reduced the T1, T4, and T12 displacement by 23.6, 15.3, and 0.5 mm, respectively. This indicates that shoulder restraint by the shoulder restraint air bag primarily acted to reduce lateral displacement at the upper region of the spine. Fig. 21. Displacement of spine (WORLDSID) (color figure available Fig. 22. Displacement of spine (THUMS) (color figure available THUMS Figure 22 compares the lateral displacement of the THUMS spine (T1, T4, and T12) with and without a shoulder restraint. The comparison was made at 51 ms, at which the maximum spinal displacement was reached in the case without the shoulder restraint air bag. The pink bars show the measures without the shoulder restraint air bag and the blue bars show the measures with the shoulder restraint air bag. Adopting the shoulder restraint air bag reduced the T1, T4, and T12 displacement by 20.4, 17.7, and 6.8 mm, respectively. This indicates that shoulder restraint by the shoulder restraint air bag primarily acted to reduce lateral displacement at the upper region of the spine. The shear forces generated at T12 were 400 N without shoulder air bag and 500 N with shoulder air bag. These force levels were relatively lower than the spine failure load between 1400 and 2100 N described by Yoganandan et al. (1988). Comparison of Shoulder and Chest Injury Measures WorldSID Figure 23 compares the shoulder and chest injury measures with and without a shoulder restraint air bag. The pink bars show the measures without the shoulder restraint air bag and the blue bars show the measures with the shoulder restraint air bag. When the shoulder restraint air bag was present, the force applied to the shoulder decreased from approximately 11 kn to approximately 2 kn in accordance with the reduction in maximum shoulder external force shown in Figure 21. In addition, although the upper thorax rib deflection decreased from approximately 55 mm to approximately 40 mm in accordance with the reduction in maximum chest force shown in Figure 22, there was virtually no change in the deflection of the middle thorax and lower thorax rib. THUMS Figure 24 compares the shoulder and chest injury measures with and without a shoulder restraint air bag. Deflection ratio (%) was obtained from the ratio between the original and measured reduction in distance from a node on the outermost side of each rib to a virtual centerline connecting the sternum and the spine. The pink bars show the measures without the

10 72 Tanaka et al. Fig. 25. Bony fracture (color figure available Fig. 23. Injury of WorldSID (color figure available shoulder restraint air bag and the blue bars show the measures with the shoulder restraint air bag. Figure 25 shows the results of the bony fracture in THUMS indicated by pink Xs for the case without the shoulder restraint air bag. The shoulder restraint air bag prevents clavicle fracture in accordance with the reduction in maximum shoulder force shown in Figure 23. In addition, although the deflection ratio from the 1st to the 10th ribs at the chest decreased in accordance with the reduction in maximum chest external force shown in Figure 24, there was an increase in chest rib deflection ratio at and below the 11th rib. The total number of rib fractures was mitigated from 4to0. Figure 26 shows contour diagrams of principal strain on the left lung as viewed from outside and inside the vehicle. Over the entire lung, the range of principal strain of 20 percent or more decreased when the shoulder restraint air bag was present. The proportion of shell elements indicating a principal strain of 20 percent or more over the entire lung area was approximately 16 percent without the shoulder restraint air bag. This decreased to approximately 5 percent with the shoulder restraint air bag. Therefore, shoulder restraint by the shoulder restraint air bag reduced 11 percent the size of the area around the left lung with a principal strain of 20 percent or more. Fig. 24. Rib deflection ratio (color figure available Fig. 26. Comparison of principal strain of lung (color figure available

11 Oblique Side Impacts 73 Fig. 27. Relationship of position between thorax rib and arm (color figure available Influence of External Force to Shoulder on Chest Injury WorldSID This section examines the influence of the external force applied to the shoulder on chest injury estimated by World- SID. The seating position of WorldSID assumes a driving posture. As a result, the arms are set at an angle to the coronal plane. Therefore, when viewed from the side, only parts of the arms overlap the chest. This section describes an example in which the driving posture contains a mixture of overlapping Fig. 29. Relationship between upper thorax rib deflection and external force applied to upper thorax rib and arm (with shoulder restraint air bag) (color figure available and nonoverlapping portions between the arms and the upper thorax rib has the highest injury value (Figure 23). Figure 27 shows the relationship between the side-view positions of the arm, shoulder, and upper thorax rib when the shoulder was restrained by the shoulder restraint air bag. In this case, the half of the upper thorax rib to the front of the dummy was overlapped by the arm. The graphs in Figures 28 and 29 show the upper thorax deflection waveform and external force applied to the upper thorax rib. The external force applied to the upper thorax rib includes the force applied to the chest through the arm and other forces (such as the force from the shoulder restraint air bag, torso air bag, and door trim). It was found that in both cases with and without the shoulder restraint air bag, the rib deflection and the external force to the rib showed similar wave forms and the peak external force to the rib induced the peak rib deflection. Therefore, with the arm interference to the upper rib, when external force was applied to the shoulder by an air bag or the trim, upper thorax rib deflection was incurred by external force from the arm. Fig. 28. Relationship between upper thorax rib deflection and external force applied to upper thorax rib and arm (without shoulder restraint air bag) (color figure available THUMS This section examines the influence of the external force applied to the shoulder on chest injury estimated by THUMS. In a driving posture, the shoulder parts such as the muscles, soft tissues, and the scapula overlap the first through seventh

12 74 Tanaka et al. Fig. 30. Transmission of external force to third rib (color figure available Fig. 32. Relationship between shoulder external force and chest injury (color figure available Discussion ribs (Figure 30). Therefore impact force applied to shoulder during the collision was transmitted to the ribs through these parts. Figure 31 compares the time history curves of external force applied to the THUMS shoulder with and without the shoulder restraint air bag. The pink line shows the force without the shoulder restraint air bag and the blue line shows the force with the shoulder restraint air bag. Neither rib nor clavicle fractured with the shoulder restraint air bag. The X mark indicates the timings of third rib fracture, which coincided with the peak external force applied to the shoulder. It was estimated that the shoulder force induced the rib fracture. Shoulder force affects the kinematics of scapula during side impact. Smith et al. (2011) reported that excessive shoulder force induced rib fractures caused by the scapula. This rib fracture mechanism is explained in Appendix A (see online supplement). To help reduce injury risk to the shoulder and the chest, excessive load input should not be applied to the shoulder. Relationship Between External Force to Shoulder and Chest Injury The relationship between the external force applied to the shoulder and the chest injury was examined. In this article, the external force was defined in elimination of bottoming force to the door trim to clarify the relationship between the shoulder air bag restraint force and injury. A parametric study was conducted on a total of 10 cases with different values for the diameter of the vent hole in the shoulder restraint air bag as shown in Figure 11. The relationship between the external force applied to the shoulder and the chest injury is shown for WorldSID in Figure 32 and for THUMS in Figure 33. The vertical axis of the plot for WorldSID indicates the maximum deflection of the upper thorax rib, and the plot for THUMS indicates the maximum deflection ratio of the third rib and the occurrence of fracture without the shoulder restraint air bag. In WorldSID, the chest deflection decreased as the external force to the shoulder increased up to 3 kn but increased beyond that. In THUMS, the thir rib fractured when the external force to the shoulder was lower than 1.8 kn but not when it was higher. The rib deflection ratio decreased with the external force from the shoulder restraint air bag to the shoulder up to 3.7 kn but bega to increase beyond that. It is speculated that one reason for the difference between World- SID and THUMS in optimum external air bag force is the Fig. 31. External force applied to shoulder (color figure available Fig. 33. Relationship between shoulder external force and chest injury (color figure available

13 Oblique Side Impacts 75 Fig. 34. Relationship between shoulder external force and chest injury of elderly occupant (70 years old) (color figure available difference in the contact area of the shoulder described in Figure 13. The contact area directly affects the external force in air bag restraint; however, in both WorldSID and THUMS, the chest injury risk decreased with the external force up to a certain level (3 to 4 kn) but began to increase beyond that. This suggests that the shoulder restraint air bag helps to decrease the chest injury risk, but the magnitude of force may need to be limited. Next, we simulated the same cases assuming an elderly occupant (70 years old). The material property of the ribs was changed as suggested by Watanabe et al. (2012). The equivalent stress and equivalent plastic strain of bone properties were approximately 86 percent lower compared to people in their 30s and 40s. Figure 34 shows the relationship between the external force applied to the shoulder and chest injury risk. The number of rib fractures was used as an index of chest injury risk because rib fractures were predicted in all cases. Similar to the previous cases, the number of rib fractures decreased with the external force up to a certain level but began to increase beyond that. In the least fracture case, the magnitude of external force was approximately 3.0 kn, which was close to that in WorldSID. The clavicle fractured when the external force applied to the shoulder reached approximately 4.3 kn. Clavicle fracture observed in this study was the case without shoulder air bag (Figure 29) and in the case of the shoulder air bag restraint force 4.3 kn was applied to an elderly occupant (Figure 34). A shoulder restraint air bag is effective for reducing the risk of clavicle fracture. Fig. 35. Relationship between external force to shoulder of THUMS and shoulder rib deflection of WorldSID (color figure available a minimum third rib deflection ratio was 3.7 kn in THUMS, corresponding to a shoulder rib deflection of 60 mm and a shoulder load of 2.2 kn in WorldSID. As for the elderly cases in THUMS, the external force with the minimum number of rib fractures was approximately 2.8 kn, corresponding to a shoulder rib deflection of 56 mm and a shoulder load of 2.1 kn in WorldSID. Further research is necessary in this area considering the occupant physique; skeletal geometry; posture, including arm position; and so on. Limitations The results obtained in this study were based on a specific impact condition assuming that a passenger vehicle strikes a rigid pole at 32 km/h with an impact angle of 75.Inanactualaccident, the occupant s physique and posture could vary greatly from this condition. This research was conducted in limited conditions; however, the authors believe that this analytical study will help us to understand the external force to shoulder in helping to mitigate chest injury in a pole side impact. The abdominal deflection of the WorldSID in FE was greater than that of the test. The door trim was modeled of an elasto-plastic shell with damage to try to simulate the vehicle tested for the model validation of occupant injury response. The contact location of the door trim and the WorldSID abdomen was close to the pole. Consequent phenomenon in the door trim was complex deformation. Further modeling study is required to simulate the contact between the complex plastic material and the occupant. Effect of Shoulder Air Bag Restraint Force on WorldSID/THUMS Chest Injury Measures The relationship between the external force applied to the shoulder in THUMS and the shoulder deflection of WorldSID in Figure 35 and the shoulder load of WorldSID in Figure 36. The specifications of the shoulder restraint air bags were exactly the same in the 2 cases. In the original THUMS case, the rib did not fracture when the external force applied to the shoulder was greater than 1.8 kn (including the greatest case of 6.3 kn). It is speculated that constraint of the shoulder at a force of 1.8 kn or greater would help reduce chest injury risk in vehicle side collisions. The shoulder air bag force with Fig. 36. Relationship between external force to shoulder of THUMS and shoulder rib load of WorldSID (color figure available

14 76 Tanaka et al. Conclusions This article examined the effectiveness of a shoulder restraint for reducing chest injury. The following points were identified: A shoulder restraint using the air bag helps reduce chest injury risk by lowering the maximum magnitude of external force to the shoulder and chest. The magnitude of external force applied to shoulder may need to be limited considering the force transmission mechanism in THUMS and WorldSID. The force was transmitted to the first through seventh ribs through the muscle, soft tissue, and scapula in THUMS. The force was transmitted to the upper thorax rib through the arm in WorldSID. To help reduce rib fracture risk in the THUMS, the shoulder restraint air bag was expected to generate a force of 3.7 kn with a minimum rib deflection ratio corresponding to a shoulder rib deflection of 60 mm and a shoulder load of 2.2 kn in WorldSID. The shoulder restraint air bag was effective in helping to reduce clavicle fracture risk in the THUMS. Excessive shoulder restraint force over 4.3 kn increased the clavicle fracture risk for an elderly occupant in this limited study. References Compiegne S, Yves C, Thierry Q, Jean PV. Non-injurious impact response of the human shoulder three-dimensional analysis of kinematics and determination of injury threshold. Stapp Car Crash J. 2004;48:1 35. DYNAmore GmbH. WorldSID 50th User s Manual. Release 1.0 for Model ver Stuttgart-Vaihingen: Dynamore/Stuttgart Industries. Fukushima S, Kumagai K, Yasuki T. Development of finite element models of restraint system for injury analysis in side impact. Paper presented at: 10th LS-DYNA Users Conference; Dearborn, MI, June 8 10, Haenchen D, Schwarz T, Thomas G, Zobel R. Feasible steps towards improved crash compatibility. Paper presented at: World Congress SAE 2004; Detroit, MI, March 8 11, Hayashi S, Yasuki T, Kitagawa Y. Occupant kinematics and estimated effectiveness of side airbag in pole side impacts using a human FE model with internal organs. Stapp Car Crash J. 2008;52: International Organisation for Standardisation DIS DRAFT Ver 5.4. Geneva, Switzerland: ISO. Kojima S, Yasuki T, Oono K. Application of shell honeycomb model to IIHS MDB model. Paper presented at: 6th LS-DYNA User s Conference; Gothenburg, Sweden, May 29, Otte D, Haasper C, Eis V, Schaefer R. Characteristics of pole impacts to side of passenger cars in European traffic accidents and assessment of injury mechanism analysis of German and UK in-depth data. Stapp Car Crash J. 2008;52: Petitjean A, Trosseille X, Petit P, Irwin A, Hassen J, Praxl N. Injury risk curves for the WorldSID 50th male dummy. Stapp Car Crash J. 2009;53: Sherer RB, Akiyama KA, Tylko S, Hartlieb M, Harigae T. World- SID production dummy biomechanical responses. Paper presented at: World Congress SAE 2009; Detroit, MI, April 20 23, Smith DR, Hayashi S, Yasuki T, Kitagawa Y. A Study of Driver Injury Mechanism in High Speed Lateral Impacts of Stock Car Auto Racing Using a Human Body FE Model. Detroit, MI: SAE International; SAE Paper Number Toyota Motor Corporation. THUMS User s Manual. Ver University of Michigan Transportation Research Institute (UMTRI). Anthropometry of Motor Vehicle Occupants Available at: Viano DC. Biomechanical responses and injuries in blunt lateral impact. Stapp Car Crash J. 1989;33: 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 J. 2012;56: Watanabe R, Miyazaki H, Kitagawa Y, Yasuki Y. Research of collision speed dependency of pedestrian head and chest injuries using human FE model (THUMS version 4). Paper presented at: 22th ESV Conference; June 13 16, 2011; Washington, DC. Yoganandan N, Pinter F, Sances JA, et al. Biomechanical Investigations of the Human Thoracolumbar Spine. Detroit, MI: SAE International; SAE Paper Number