Appendix A - Figures. A 2 Chestband contours for the twelve cases (a) A-pillar (b) B-pillar (c) C-pillar

Transcription

1 Appendix A - Figures A 1 GHBMC model at t=0 ms (a) with no belt tightening (b) with belt tightening; seat belt hidden for displacement contour visibility A 2 Chestband contours for the twelve cases (a) A-pillar (b) B-pillar (c) C-pillar

2 A 3 Plastic strain concentrated on the right and left quadrants of the rib cage A 4 Rib cage loading due to lap belt force (blue arrow), and torso inertia (red arrow) A 5 Pelvis deformation due to lap belt loading

3 Appendix B Tables Method Parameters Values Description Corridor K 2 Transitional order of function between ratings of 1 and 0 G_1 0.5 Weighting factor of the corridor method A_ Width of the inner corridor B_0 0.5 Width of the outer corridor A_SIGMA 0 Parameter to widen the inner corridor B_SIGMA 0 Parameter to widen the outer corridor Cross correlation D_MIN 0.01 Minimum interval of evaluation D_MAX 0.12 Maximum interval of evaluation INT_MIN 0.8 Minimum interval overlap K_V 10 Transitional order of function between ratings of 1 and 0 for progression K_G 1 Transitional order of function between ratings of 1 and 0 for size K_P 1 Transitional order of function between ratings of 1 and 0 for phase G_V 0.5 Weighting factors of the progression rating G_G 0.25 Weighting factors of the size rating G_P 0.25 Weighting factors of the phase shift rating G_2 0.5 Weighting factors of the cross correlation method B 1 CORA parameters used in this study

4 Authors Specimen type Organ Ultimate Strain (%) J. G. Snedeker et al. (2005) Human Kidney 27 to 34 Untaroiu et al. (2013) Human Liver 22 to 37 Kemper et al. (2012) Human Spleen 17 to 26 Pedro et al. (2011) Human Bladder 60 to 80 Yamada, 1969 Dogs Lungs 15 B 2 Source of threshold values used in the current study

5 Appendix C - Text CORA was used to quantify the goodness-of-fit between simulation and experimental responses. The ratings in CORA range from 0 to 1, where 0 representing poor correlation, and 1 representing perfect match. CORA uses two methods that are fundamentally very different to quantify the degree of correlations between the responses. The first method is the corridor method that calculates the correlation between the responses using user-defined corridors (inner and outer). In this study, the inner corridors were calculated for 5 percent of the peak mean trace, whereas the outer corridor was calculated for 50 percent of the peak mean trace with default parameters (Table B1). The second method is the cross-correlation method that is more involved and complex, for brevity, only a brief description is presented here. A more detailed description can be found in Gehre, Gades et al. (2009). This method quantifies the progression, time-shift, and size of the responses, using three respective metrics. All the three metrics are calculated using the time-shifted data for the maximum cross correlation value. Weighted sum of these three metrics gives the rating for correlation method. Equally weighted sum of the corridor, and cross-correlation methods give the overall correlation rating. C 1 CORA description In the PMHS experiments used for FE-HBM validation, to simulate the pretensioning effect, the shoulder belt was tightened prior to the experiments. To mimic a realistic tightening effect in the simulations, a workflow with three phases was performed. To realistically simulate the placement of the PMHS on the rigid seat, the GHBMC model was settled using acceleration due to gravity. The complete settling of the GHBMC model was confirmed by verifying the absence of transient component in the contact force between the GHBMC model and the seat pan. Once the model was settled on the rigid seat, a prescribed motion was applied to the end of the shoulder belt in the A- P direction. The belt tightening phase simulation was performed for a predetermined period and the deformed state of the model was extracted and used as the initial setup for far-side impact simulations. The initial velocity was applied to the whole model (8.3 m/s) and the deceleration pulse was applied to the rigid seat. All simulations were solved using LS-DYNA MPP version 7.0 solver on a Linux RHEL 5.4 computational cluster, using 64 cores. C 2 FE-HBM validation setup description