SAFETY ENHANCED INNOVATIONS FOR OLDER ROAD USERS EUROPEAN COMMISSION EIGHTH FRAMEWORK PROGRAMME HORIZON 2020 GA No. 636136 The research leading to the results of this work has received funding from the European Union's Horizon 2020 research and innovation programme under grant agreement No 636136. Deliverable No. D3.2b Deliverable Title Updated Impactor Test and Validation Report Dissemination level Public Written by Zander, Oliver BASt Ott, Julian Lundgren, Christer Autoliv Fornells, Alba IDIADA Luera, Andrea FCA Checked by David Hynd TRL 26/05/2018 Mark Burleigh Humanetics 26/05/2018 Approved by Wisch, Marcus BASt 26/05/2018 Issue date 29/05/2018
107
Embed
SAFETY ENHANCED INNOVATIONS FOR OLDER ROAD USERS … · Issue date . 29/05/2018 : Deliverable 3.2b Page | 2 out of 107 ... impactor simulations vs. actual vehicle models and generic
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
SAFETY ENHANCED INNOVATIONS FOR OLDER ROAD USERS
EUROPEAN COMMISSION EIGHTH FRAMEWORK PROGRAMME
HORIZON 2020 GA No. 636136
The research leading to the results of this work has received funding
from the European Union's Horizon 2020 research and innovation
programme under grant agreement No 636136.
Deliverable No. D3.2b
Deliverable Title Updated Impactor Test and Validation Report
Dissemination level Public
Written by
Zander, Oliver BASt
Ott, Julian
Lundgren, Christer Autoliv
Fornells, Alba IDIADA
Luera, Andrea FCA
Checked by David Hynd TRL 26/05/2018
Mark Burleigh Humanetics 26/05/2018
Approved by Wisch, Marcus BASt 26/05/2018
Issue date 29/05/2018
Deliverable 3.2b
Page | 2 out of 107
EXECUTIVE SUMMARY
The demographic change european society is facing during the next decades will be
challenging also for passive vehicle safety. The external road user safety branch of
the HORIZON 2020 SENIORS project addresses special safety needs in particular of
the elderly by defining equivalent safety requirements to passenger cars within test
and assessment procedures, alongside with a provision of new and revised test
tools, towards an appropriate assessment of the vehicle protection potential, also
taking into account the ongoing changes in injury patterns of vulnerable road users
since the introduction of consumer test programmes and regulative requirements.
This report resumes the results of baseline pedestrian simulations with human body
models and impactor models against generic test rigs. For that purpose, subsequent
to the work reported about in Deliverable D2.5b (Zander et al., 2017), various
impactor simulations vs. actual vehicle models and a generic SAE Buck and its
derivatives (representing a Sedan, an SUV and a Van/MPV frontend) have been
carried out and compared to the results from human body model simulations against
identical frontends. The Deliverable compares kinematics, time histories as well as
peak loadings and identifies possible correlations between the loadings to the human
body model and the impactor models. The results will be used to establish test and
assessment procedures for vulnerable road users in a later project stage and to be
evaluated by means of physical component and full-scale tests.
Deliverable 3.2b
Page | 3 out of 107
Following participants contributed to this deliverable report: Partner Representative Chapters BASt Oliver Zander 1-4 BASt Julian Ott 3 Autoliv Christer Lundgren 3.2 IDIADA Alba Fornells 3.2, 4.2 FCA Andrea Luera 3.2
1.1 The EU Project SENIORS ................................................................................................ 5
1.2 Background and Objectives of this Deliverable .............................................................. 5 2 Simulation programme .............................................................................................. 6
3.2.2 Actual vehicles ........................................................................................ 82 3.2.2.1 SUV ...................................................................................................................................... 82
4 Summary and conclusions........................................................................................ 91
4.1 Conclusions of validations and impact on SENIORS project .......................................... 91
4.2 Transfer of results ...................................................................................................... 93 Glossary ......................................................................................................................... 99 References ..................................................................................................................... 100 Acknowledgements ....................................................................................................... 102 Disclaimer...................................................................................................................... 102 APPENDIX A ................................................................................................................... 103 Diagrams ....................................................................................................................... 103
Deliverable 3.2b
Page | 5 out of 107
1 INTRODUCTION
1.1 THE EU PROJECT SENIORS Because society is aging demographically and obesity is becoming more prevalent,
the SENIORS (Safety ENhanced Innovations for Older Road userS) project aims to
improve the safe mobility of the elderly, and overweight/obese persons, using an
integrated approach that covers the main modes of transport as well as the specific
requirements of this vulnerable road user group.
This project primarily investigates and assesses the injury reduction in road traffic
crashes that can be achieved through innovative and suitable tools, test and
assessment procedures, as well as safety systems in the area of passive vehicle
safety. The goal is to reduce the numbers of fatally and seriously injured older road
users for both major groups: car occupants and external road users (pedestrians,
cyclists, e-bike riders).
Implemented in a project structure, the SENIORS project consists of four technical
Work Packages (WP1 – WP4) which interact and will provide the substantial
knowledge needed throughout the project. These WPs are:
WP1: Accidentology and behaviour of elderly in road traffic
WP2: Biomechanics
WP3: Test tool development
WP4: Current protection and impact of new safety systems
In addition, there is one Work Package assigned for the Dissemination and
Exploitation (WP5) as well as one Work Package for the Project Management (WP6).
1.2 BACKGROUND AND OBJECTIVES OF THIS DELIVERABLE The demographic change European society is facing during the next decades will be
challenging also for passive vehicle safety. The external road user safety branch of
the HORIZON 2020 SENIORS project will address special safety needs in particular
of the elderly by defining equivalent safety requirements to passenger cars within test
and assessment procedures, alongside with a provision of new and revised test
tools, towards an appropriate assessment of the vehicle protection potential.
Deliverable 3.2b
Page | 6 out of 107
Deliverable D3.2b of the SENIORS project resumes the results of baseline
pedestrian simulations with human body models and impactor models against
generic test rigs. For that purpose, subsequent to the work reported about in
Deliverable D2.5b, various impactor simulations vs. actual vehicle models and a
generic SAE Buck and its derivatives (representing a Sedan, an SUV and a
Van/MPV frontend) are carried out and compared to the results from human body
model simulations against identical frontends. Impact kinematics, time histories as
well as peak loadings are compared and possible correlations between the loadings
to the human body model and the impactor models defined.
2 SIMULATION PROGRAMME
2.1 AIM
The simulation programme reported about in this Deliverable is intended as
validation of impactor results formerly obtained against generic vehicle frontends vs.
actual vehicle models on the one hand, and as a further investigation of possible
correlations between human body models simulations and impactor simulations, as
depicted in Figure 1 on the other hand. These correlations would then go into the
assessment procedure further described in SENIORS Deliverables D4.1(b) (Zander
et al., 2018-2) and D4.2(b) (Zander et al., 2018-3).
Deliverable 3.2b
Page | 7 out of 107
Figure 1: Context of impactor validations within overall pedestrian flowchart, example lower extremities.
2.2 SIMULATION MATRIX
Subsequent to baseline simulations with the human body model THUMS v4 (Total
HUman Model for Safety), FlexPLI (Flexible Pedestrian Legform Impactor), FlexPLI-
UBM (Upper Body Mass), HNI (Head Neck Impactor) and TIPT (Thorax Injury
Prediction Tool) vs. a generic test rig and generic vehicle frontends, as reported in
Deliverable D2.5b (Zander et al., 2018-1), a variety of simulations with the mentioned
impactors were carried out against actual vehicles, the SAE Buck and its derivatives,
representing a Sedan, an SUV and a Van/MPV. Since baseline simulations with the
HNI did not demonstrate a potential benefit in terms of an improved kinematic
behaviour nor in regarding the impactor readings in simulations against the SAE
Buck, this approach was not followed further during the project. Thus, subsequent
work was focused on the refinement of the FlexPLI with applied upper body mass as
well as the further development of a prediction tool for thoracic injuries of vulnerable
Deliverable 3.2b
Page | 8 out of 107
road users. An overview of all validation simulations regarding the FlexPLI-UBM
development including ambient conditions is illustrated in Table 1.
Table 1: Context of impactor validations within overall pedestrian flowchart.
The maximum bending moments and elongations during simulations against the
actual Van/MPV are summarised in Figure 37:
Figure 37: Peak bending moments and ligament elongations, actual Van/MPV.
In most cases, the peak loadings on the lower extremities are overpredicted by the
FlexPLI with upper body mass; however, the high maximum MCL elongations of the
HBM during impacts against the vehicle centreline as well as the end of the bumper
beam are much better reflected by the UBM versions of the impactor, as can also be
exemplarily seen for the good qualitative correlation in Figure 38 for vehicle
centreline.
Deliverable 3.2b
Page | 36 out of 107
Figure 38: Time histories of MCL elongation during simulations at vehicle centreline against actual Van/MPV.
Besides, also the peak femur bending moments at the end of the beam (RHS) are
better represented by FlexPLI-UBMrigid and FlexPLI-UBMrubber.
The improved qualitative correlation between HBM and FlexPLI at the end of the
bumper beam when using the upper body mass is also demonstrated by the impact
kinematics, as illustrated for the time of maximum loadings of THUMSv4 and the
FlexPLI-UBM:
Deliverable 3.2b
Page | 37 out of 107
Figure 39: Impactor and HBM kinematics during impact against actual MPV at EoB for time of maximum HBM loading (upper) and the time of maximum FlexPLI-UBM loading (lower) – side view.
It can be seen that also in terms of rotation of the lower extremities the FlexPLI-
UBMrigid as well as the FlexPLI-UBMrubber getting much closer to the HBM than the
FlexPLI Baseline.
Deliverable 3.2b
Page | 38 out of 107
Figure 40: Impactor and HBM kinematics during impact against actual MPV at EoB for time of maximum HBM loading (upper) and the time of maximum FlexPLI-UBM loading (lower) – top view.
A comparison of the FlexPLI with and without applied UBM and the HBM shows that
the rotation of the human lower extremities is overpredicted by the FlexPLI Baseline
but in its extent well represented by the FlexPLI-UBM, though with a certain time lag,
as demonstrated in Figure 41:
Figure 41: Comparison of Impactor and HBM rotation during impact against actual MPV at EoB.
Deliverable 3.2b
Page | 39 out of 107
Figure 42 summarises the peak results during simulations against the SUV vehicle
model:
Figure 42: Peak bending moments and ligament elongations, actual SUV.
In all cases, the rigidly attached UBM signs for an overprediction of femur bending
moments, while the peak bending moments of the HBM are best represented by the
FlexPLI-UBMrubber. This is also the case for the waveforms, see Figure 43:
Figure 43: Time histories of Femur-1 bending moment (HBM: My) during simulations at vehicle centreline against actual SUV.
Deliverable 3.2b
Page | 40 out of 107
Regarding the ligament elongations, both FlexPLI versions with UBM well represent
the human loads, in many cases also with respect to the shapes of the waveforms.
However, the extraordinary high peak value of MCL at SUV centreline was not
observed during the simulations with FlexPLI and its derivatives.
Peak bending moments and ligament elongations during simulations against the
Compact Car representative are depicted in Figure 44:
Figure 44: Peak bending moments and ligament elongations, actual Compact Car.
In particular, the peak femur and tibia bending moments as well as the MCL
elongation at vehicle centreline are well represented by FlexPLI-UBMrigid and
FlexPLI-UBMrubber. This is also demonstrated by the time history curves, as
exemplarily shown for Femur-1 in Figure 45:
Deliverable 3.2b
Page | 41 out of 107
Figure 45: Time histories of Femur-1 bending moment (HBM: My) during simulations at vehicle centreline against actual Compact Car.
Towards the end of the bumper beam with more angled surfaces, the upper body
masses seem to overpredict the femur loads and to underpredict the tibia loads.
On the other hand, duration and shape of the HBM waveforms are still much better
reflected by the UBM impactors, as shown for the lowermost femur segment in
Figure 46:
Deliverable 3.2b
Page | 42 out of 107
Figure 46: Time histories of Femur-1 bending moment (HBM: My) during simulations at EoB against actual Compact Car.
Figure 47 summarises the peak bending moment and maximum elongation results
during the simulations against the Limousine:
Figure 47: Peak bending moments and ligament elongations, actual Limousine.
The diagrams demonstrate in particular the benefit of FlexPLI with applied UBM on
the peak bending moments as well as MCL elongations in comparison to the FlexPLI
Baseline. The superior behaviour of FlexPLI-UBMrigid and FlexPLI-UBMrubber
compared to the Baseline impactor is confirmed by the time history curves, as
exemplarily shown in Figure 48 for Femur-2 and in Figure 49 for MCL:
Deliverable 3.2b
Page | 43 out of 107
Figure 48: Time histories of Femur-2 bending moment (HBM: My) during simulations at vehicle centreline against actual Limousine.
Figure 49: Time histories of MCL elongation during simulations at vehicle centreline against actual Limousine.
A comparison of the simulation results against the different versions of the SUV Buck
and four actual vehicle models reveals the significantly superior behaviour of the
FlexPLI-UBMrigid and FlexPLI-UBMrubber over the FlexPLI Baseline in terms of
Deliverable 3.2b
Page | 44 out of 107
kinematics and biofidelity, with further advantages of the FlexPLI-UBMrubber such as
the ability of simulating the hip rotation and time lag (Figure 50), a sometimes later
UBM interaction with vehicle front, leading to more humanlike results, and a
sometimes slightly more realistic shape of the waveforms. This however was only
partially reflected by the peak loadings on femur, tibia, and knee ligaments.
Subsequent correlation studies will further investigate this phenomenon.
Figure 50: Impact kinematics during simulations against the generic test rig (Zander et al., 2017).
As described in the previous chapter, the application of a pedestrian torso mass to
the flexible pedestrian legform impactor FlexPLI contributed in most cases to a
significant improvement of the kinematics and impact biofidelity when being
compared to the human body model THUMS4 under identical loads. Nonetheless,
this improvement was not always reflected in peak femur and tibia bending moments
and knee elongations closer to those of THUMS.
Deliverable 3.2b
Page | 45 out of 107
3.1.3.2.1 All impacts
Figure 51 summarises the quantitative correlations (peak values) for the maximum
femur and tibia bending moments between THUMSv4 and FlexPLI Baseline,
FlexPLI-UBMrigid and FlexPLI-UBMrubber calculated from the simulations carried out
against the SAE Buck and its derivatives and the four actual vehicle models
representing a Van/MPV, an SUV, a Compact Car and a Limousine. It has to be
noted that while for the two dimensional SAE Buck frontend only one impact point at
vehicle centreline was chosen, all the actual vehicle models were additionally
impacted at both ends of the bumper beam for investigating the behaviour of the
impactor at angled surfaces.
Figure 51: Quantitative correlations between THUMSv4 and FlexPLI and its derivatives – maximum femur and tibia bending moments (all impacts).
A comparison of the coefficient of determination R2 for linear regression surprisingly
shows the best correlation with HBM in the femur area for FlexPLI Baseline, followed
by FlexPLI-UBMrubber. The correlation between HBM and FlexPLI-UBMrigid is poor.
With regards to the tibia area all correlations in terms of peak values are
unsatisfactory, with the FlexPLI-UBMrubber correlating best. This for the FlexPLI
Baseline unexpected result is inconsistent to the good biofidelity and injury
assessment ability attributed to its tibia area, see Figure 52. The additional torso
mass is meant towards an improvement of biofidelity mainly in the femur area;
however, as demonstrated in Figure 51, the biofidelity of the baseline impactor is
best for femur, which is not in line with the biofidelity determined during the
development of the FlexPLI.
Deliverable 3.2b
Page | 46 out of 107
Figure 52: Quantitative correlations between Human FE model and FlexPLI– maximum tibia bending moments (JASIC, 2016).
Figure 53 illustrates the quantitative correlations of maximum bending moments in
the particular femur segments between THUMSv4 and the FlexPLI and its
derivatives:
Figure 53: Quantitative correlations between THUMSv4 and FlexPLI and its derivatives – bending moments of femur segments (all impacts).
While the FlexPLI Baseline maintains a similar correlation for all femur segments (R²
between 0.44 and 0.57), the FlexPLI-UBMrigid and FlexPLI-UBMrubber correlate the
better with the HBM the closer to the knee. At the lowermost femur segment, FlexPLI
Baseline correlates with the best coefficient of determination (R²=0.56), followed by
FlexPLI-UBMrubber (R²=0.49) and FlexPLI-UBMrigid (R²=0.42). Altogether, in terms of
peak bending moments, an additional UBM did not lead to a significantly better
correlation with the HBM.
Deliverable 3.2b
Page | 47 out of 107
The quantitative correlation of the peak tibia bending moments in the different areas
of the lower leg only partly confirm the observation from the maximum moments over
the entire length of the tibia:
Figure 54: Quantitative correlations between THUMSv4 and FlexPLI and its derivatives – bending moments of tibia segments (all impacts).
At tibia-1, the quantitative correlation in terms of maxima between HBM and FlexPLI
is poor. Since most of the maximum tibia bending moments are acquired close to the
knee, the upmost tibia bending moment signs responsible for the overall poor
correlation as depicted in Figure 51. On the other hand, correlation gets better with
increased distance of the particular impactor strain gauge from the knee. Altogether,
the best correlation can be found for FlexPLI-UBMrubber, followed by FlexPLI-UBMrigid.
The quantitative correlation of maximum knee elongations between HBM and
FlexPLI, FlexPLI-UBMrigid and FlexPLI-UBMrubber is summarised in Figure 55:
Deliverable 3.2b
Page | 48 out of 107
Figure 55: Quantitative correlations between THUMSv4 and FlexPLI and its derivatives – elongations of knee ligaments (all impacts).
While the correlation between THUMSv4 and the FlexPLI, the FlexPLI-UBMrigid and
FlexPLI-UBMrubber is very poor for the cruciate ligament elongations, correlation of
MCL is best for the FlexPLI (R²=0.33), followed by FlexPLI-UBMrigid (R²=0.24) and
FlexPLI-UBMrubber (R²=0.22). As for the maximum tibia bending moments, the ACL
correlations between the human FE model and the FlexPLI Baseline was found much
better during the development of the FlexPLI, see Figure 56. On the other hand, the
comparatively unsatisfactory MCL correlation during FlexPLI development was
confirmed.
Figure 56: Quantitative correlations between Human FE model and FlexPLI – maximum ligament elongations (JASIC, 2016).
Altogether, the good correlations between HBM and FlexPLI-UBM in terms of
kinematics and time histories were not confirmed under consideration of correlating
maximum loadings in all impacts. This effect needs to be considered during the
Deliverable 3.2b
Page | 49 out of 107
subsequent establishment of impactor thresholds for the FlexPLI-UBM in the
framework of new test and assessment procedures for implementation within
regulatory and consumer programmes.
3.1.3.2.2 Centreline impacts Figure 57 shows the quantitative correlations between THUMSv4 and FlexPLI,
FlexPLI-UBMrigid and FlexPLI-UBMrubber for all impacts against vehicle centreline,
only:
Figure 57: Quantitative correlations between THUMSv4 and FlexPLI and its derivatives – maximum femur and tibia bending moments (centreline impacts).
Neglecting the simulations against angled surfaces significantly improves the
coefficient of determination for FlexPLI Baseline in the femur area. For FlexPLI with
UBM the evaluation of perpendicular impacts only has no positive effect. It can be
concluded that test results of the FlexPLI Baseline are sensitive to the inclination of
the vehicle frontend, with a lower effect on the UBM derivatives.
The positive effect of evaluating the perpendicular impacts only on FlexPLI Baseline
femur correlation is confirmed when focusing on particular segments, see Figure 58:
Figure 58: Quantitative correlations between THUMSv4 and FlexPLI and its derivatives – bending moments of femur segments (centreline impacts).
Deliverable 3.2b
Page | 50 out of 107
For all strain gauges, the highest correlation is found between THUMSv4 and
FlexPLI Baseline, followed by FlexPLI-UBMrubber and FlexPLI-UBMrigid. The segments
confirm a significantly improved correlation for FlexPLI Baseline when eliminating the
oblique impacts on the one hand, and no further improvement for FlexPLI-UBM on
the other hand. It can be concluded the robustness of FlexPLI-UBM towards impacts
on angles surfaces while the degree of correlation of the FlexPLI Baseline depends
to a great extent on angle of the impacted surface.
In terms of the maximum tibia bending moment, no meaningful correlation can be
established under evaluation of the centreline impacts, only (see Figure 57). This
tendency can be confirmed with an evaluation of the different tibia strain gauges:
Figure 59: Quantitative correlations between THUMSv4 and FlexPLI and its derivatives – bending moments of tibia segments (centreline impacts).
The correlation of ligament elongations during impacts against vehicle centrelines are
summarised in Figure 60:
Deliverable 3.2b
Page | 51 out of 107
Figure 60: Quantitative correlations between THUMSv4 and FlexPLI and its derivatives – elongations of knee ligaments (centreline impacts).
As under consideration of all impacts (compare Figure 55), no correlation could be
established for the maximum cruciate ligament elongations. For MCL however, the
correlation between THUMSv4 and FlexPLI-UBM is significantly improved, with no
improvement for FlexPLI Baseline.
3.1.3.2.3 Vehicle categorization For a categorization of quantitative correlations of peak bending moments and
maximum ligament elongations a limited number of results was available for each
vehicle category, only. Since for the SUV category only the simulations at vehicle
centreline could be carried out with the FlexPLI-UBMrubber under the proposed impact
conditions, the linear regression always results in a coefficient of determination of 1
without providing information on the actual degree of correlation. For the Van/MPV
category only four data points per impactor are available, decreasing the quality of
correlation analysis. However, the results of both vehicle categories are plotted in the
annex to this report.
The quantitative correlations of maximum femur and tibia bending moments between
THUMSv4 and FlexPLI Baseline, FlexPLI-UBMrigid and FlexPLI-UBMrubber for impacts
against the Sedan vehicle category (Compact Car, Limousine and SAE Buck – seven
data points altogether) are depicted in Figure 61. Best correlations can be stated for
both versions of the FlexPLI-UBM in the femur area (R²=0.79 and 0.77). In the tibia
area, FlexPLI-UBMrigid correlates best with THUMSv4, followed by FlexPLI Baseline
and FlexPLI-UBMrubber. Thus, the correlation of the FlexPLI-UBM bending moments
is mostly superior to the FlexPLI Baseline when tested against Sedan vehicles.
Deliverable 3.2b
Page | 52 out of 107
Figure 61: Quantitative correlations between THUMSv4 and FlexPLI and its derivatives – maximum femur and tibia bending moments (Sedan).
A comparison by sections of the femur correlations is plotted in Figure 62, where the
very good correlation of FlexPLI-UBMrigid and FlexPLI-UBMrubber can be stressed in
particular for the two uppermost femur segments.
Figure 62: Quantitative correlations between THUMSv4 and FlexPLI and its derivatives – bending moments of femur segments (Sedan).
The maximum tibia bending moment correlation as shown in Figure 61 is mainly
driven by the peak bending moments in the upmost tibia area with the highest
loadings close to the knee, see Figure 63. For all remaining tibia sections, the peak
correlations are much better (R² between 0.72 and 0.8 for FlexPLI-UBMrigid and
between 0.66 and 0.84 for FlexPLI-UBMrubber), with the FlexPLI-UBM superior to the
FlexPLI Baseline.
Deliverable 3.2b
Page | 53 out of 107
Figure 63: Quantitative correlations between THUMSv4 and FlexPLI and its derivatives – bending moments of tibia segments (Sedan).
The correlations of the maximum knee ligament elongations for the Sedan category
are depicted in Figure 64. Altogether, best correlations can be observed for the
FlexPLI-UBMrigid at ACL and MCL (R²=0.26 and 0.21) and for the FlexPLI-UBMrubber
at MCL (R²=0.21).
Figure 64: Quantitative correlations between THUMSv4 and FlexPLI and its derivatives – elongations of knee ligaments (Sedan).
To summarise, the good qualitative correlations between the human body model
THUMSv4 and FlexPLI-UBM in terms of impact kinematics and time histories could
not be confirmed by the correlation of peak values. No relationship between this
Deliverable 3.2b
Page | 54 out of 107
phenomenon and angled impacts could be found; thus both the FlexPLI-UBMrigid as
well as the FlexPLI-UBMrubber reveal improvements compared to the FlexPLI
Baseline when tested on areas at or around the end of the bumper beam. Further
analysis of peak values results in overall good correlations between THUMSv4 and
FlexPLI-UBM for the Sedan category with a superior behaviour to the FlexPLI
Baseline. The individual number of data points of SUV and Van/MPV categories did
not further allow a reliable correlation of peak results. However, as displayed in the
annex, the results are not satisfactory. A reason thus for the altogether not always
good correlations, as particularly shown in Figure 51, can be suspected due to the
results from SUV (upper femur and tibia) impacts with unintended or premature
interactions between the upper body mass and the vehicle frontend.
As an example, the time histories of the femur bending moments during simulations
against the actual SUV at vehicle centreline are plotted in Figure 65. While no
significant difference between the waveforms of FlexPLI Baseline at the different
measurement locations can be observed, the shape and duration of the time histories
of FlexPLI-UBMrigid as well as FlexPLI-UBMrubber are getting closer to those of the
human body model when moving downwards:
Figure 65: Time histories of Femur bending moments (HBM: My) during simulations at vehicle centreline against actual SUV.
Figure 66 depicts the kinematics of the human body model THUMSv4 and the
different FlexPLI impactors at different points in time of the impact against the actual
SUV at vehicle centreline. As can be seen in the waveforms of Figure 65, both
versions of the FlexPLI with UBM have their maximum at around 30 ms after the first
contact, caused by the interaction and stop of forward motion of the upper body mass
with the leading edge of the vehicle, while the HBM reaches the maximum loadings
Deliverable 3.2b
Page | 55 out of 107
of the femur approx. 10 ms later, likely due to the impact of the human pelvis with the
upper vehicle load path. The FlexPLI-UBM maxima thus are driven by the impact of
the additional torso mass at a point in time different to the human pelvis being
loaded. It can be concluded that under modification of the UBM the femur would
reach its maximum at a later point in time and thus would correlate better with the
human femur in terms of peak loadings.
Figure 66: Impact kinematics during simulations against actual SUV at vehicle centreline.
Altogether, removing outline with premature interaction between UBM and vehicle
front is expected to contribute to a better peak correlation between HBM and
FlexPLI-UBM. Simulations against the Sedan frontends with good femur correlations
(see Figure 61) support this hypothesis. All three Sedan representatives (SAE Buck,
Deliverable 3.2b
Page | 56 out of 107
Compact Car, Limousine) interact between their upper body surrogate and the
vehicle front after having reached their maximum femur bending moments.
3.1.3.3 Vehicle rotation During component tests with the FlexPLI, an extraordinary rotation around the z-axis
of the impactor is sometimes observed when tested against angled surfaces, e.g. in
particular at the end of the test area. Since the test area is limited by either the ends
of the bumper beam or the corners of bumper, as defined by the contact points
between a vertical plane or corner gauges making an angle of 60 degrees to the
vertical longitudinal vehicle centreplane, it seems convenient for achieving a
perpendicular impact to rotate the vehicle by 30 degrees around its z axis for the
impact, as illustrated in Figure 67:
Figure 67: Vehicle rotation to compensate for FlexPLI rotation at angled surfaces.
For estimation of the additional benefit of a changed procedure at the end of the test
area, additional simulations have been carried out against the actual Van/MPV
representative at both ends of the bumper beam, using both the FlexPLI-UBMrigid and
the FlexPLI-UBMrubber.
60°60°
30°
Deliverable 3.2b
Page | 57 out of 107
Figure 68 compares the femur-3, tibia-1 and MCL time histories for the impact point
at the end of the bumper beam (LHS) with and without the vehicle being rotated by
30 degrees:
Figure 68: Simulations against actual Van/MPV at the end of the bumper beam (LHS) without (left) and with (right) rotation of the vehicle.
The time histories sometimes show a change of peak value (femur and MCL) as well
as the shape of the waveform during the time of maximum loadings; however getting
closer to the human body model results in case of the tibia bending moments only.
A comparison of impactor kinematics with and without rotation of the vehicle during
time of maximum loading is illustrated in Figure 69; kinematics after initiation of
impactor release from the vehicle front in Figure 70:
Deliverable 3.2b
Page | 58 out of 107
Figure 69: Impactor and HBM kinematics during impact against non-rotated and rotated vehicle (time of maximum loading).
w/o Vehicle Rotation
30° Vehicle Rotation
Deliverable 3.2b
Page | 59 out of 107
Figure 70: Impactor kinematics during impact against non-rotated and rotated vehicle (release of impactor).
The illustrations show that during the impact against the non-rotated vehicle the
impactor is transferred into z rotation after having reached its maximum loadings.
When impacting the vehicle rotated by 30 degrees, no additional impactor rotation
can be noted at the same point in time. It can be concluded that a perpendicular
impact can contribute to a minimization of unrealistic impactor rotation during the
impact.
The comparison of quantitative correlations for the Van/MPV shows a significant
improvement for the rotated vehicle with respect to the maximum femur and tibia
bending moment results with decreasing degree of correlation for MCL at the same
time, see Figure 71:
w/o Vehicle Rotation
30° Vehicle Rotation
Deliverable 3.2b
Page | 60 out of 107
Figure 71: Comparison of quantitative correlation during simulations against the actual Van/MPV without (left) and with (right) rotation of the vehicle.
The impactor behaviour during tests against a rotated vehicle will be further
investigated during physical tests and reported in SENIORS Deliverable D4.2(b).
R² = 0,7288
R² = 0,091R² = 0,5068
0
50
100
150
200
250
300
350
400
450
500
200 250 300 350 400
Femurmax
THUMS vs. FlexPLI
THUMS vs. FlexPLI-UBMrigid
THUMS vs. FlexPLI-UBM rubber
R² = 0,8451
R² = 0,2536
R² = 0,1785
0
50
100
150
200
250
300
350
100 150 200 250
Tibiamax
THUMS vs. FlexPLI
THUMS vs. FlexPLI-UBMrigid
THUMS vs. FlexPLI-UBM rubber
R² = 0,8622
R² = 0,4417
R² = 0,2752
0
5
10
15
20
25
30
35
40
45
15 20 25 30 35
MCL
THUMS vs. FlexPLI
THUMS vs. FlexPLI-UBMrigid
THUMS vs. FlexPLI-UBM rubber
R² = 0,7288
R² = 0,0868R² = 0,6208
0
50
100
150
200
250
300
350
400
450
500
200 250 300 350 400
Femurmax (Vehicle Rotation)
THUMS vs. FlexPLI
THUMS vs. FlexPLI-UBMrigid
THUMS vs. FlexPLI-UBM rubber
R² = 0,8451
R² = 0,3768
R² = 0,2084
0
50
100
150
200
250
300
350
100 150 200 250
Tibiamax (Vehicle Rotation)
THUMS vs. FlexPLI
THUMS vs. FlexPLI-UBMrigid
THUMS vs. FlexPLI-UBM rubber
R² = 0,8622
R² = 0,218
R² = 0,1821
0
5
10
15
20
25
30
35
40
45
15 20 25 30 35
MCL (Vehicle Rotation)
THUMS vs. FlexPLI
THUMS vs. FlexPLI-UBMrigid
THUMS vs. FlexPLI-UBM rubber
Deliverable 3.2b
Page | 61 out of 107
3.2 THORAX
3.2.1 SAE Buck
3.2.1.1 HBM (TUC THUMS) Since the following investigations were focused on injuries related to the thorax
rather than to those to the lower extremities, THUMS TUC version 2.01 was used.
The THUMS TUC V2.01 was positioned to SAE stance according to the table in
Surface Vehicle Information Report, J2782, Sep 2009, see table below, and impacted
on its right side i.e. rear leg impacted first.
Directions from SAE J2782 Issued proposed draft Sep 2009
Segment Aspect Units Axis (SAE J211-1) Target Tol TUC
V2.0.1 A Head Angle deg About Y -7 ±5 5.8 B deg About X 0 ±5 0 C Torso Angle deg About Y 83 ±5
D deg About X 0 ±5 0 E Knee Height Impact Side mm Z 505 ±10 500 F Non-Impact
Side mm Z 520 ±10 530 G Knee Bend Angle Impact Side deg Angle in XZ-
plane 164 ±5 169 H Non-Impact
Side deg Angle in XZ-plane 171 ±5 173
I Tibia Angle Impact Side deg About Y 73 ±5 75 J Non-Impact
Side deg About Y 98 ±5 100 K Femur Angle Impact Side deg About Y 89 ±5 85 L Non-Impact
Side deg About Y 107 ±5 104 M Impact Side deg About X 87 ±5 91 N Non-Impact
Side deg About X 94 ±5 91 O Knee To Knee
Width mm Y 280 ±10 236 P Heel To Heel
Distance mm X 310 ±10 310 Q mm Y 280 ±10 287 R Elbow to Elbow
Width mm Y 420 ±10 424 S Upper Arm Angle deg About X 65 ±5 86 T Upper Arm Angle deg About X 65 ±5 88 Figure 72: SAE positioning table.
The THUMS TUC model was impacted by three different buck models, SAE, MPV
and SUV buck, each one with four different velocities 20km/h, 30km/h, 40km/h and
50km/h.
TUC V2.01 SAE Stance
Deliverable 3.2b
Page | 62 out of 107
Figure 73: THUMS Impacted by SAE, MPV and SUV buck.
Tracking points on THUMS TUC V2.01 that were recorded during simulations were
head COG, neck (C1 and C7), thorax (T1 and T12), and pelvis. Rib 4, 6 and 8 lateral
deflections were also recorded with implemented spring elements. In this section only
thorax results will be presented.
Table 6: Thorax impact timing was taken when complete thorax side is in contact with buck.
Thorax impact velocity relative to the buck is dependent on both buck impact velocity
and geometry. The SAE buck impacts THUMS lower on the legs compared to the
other two buck geometries and gives the highest thorax rotational velocity and as a
consequence the highest impact velocity relative to the buck.
T1 and T12 impact velocities relative to the buck can be seen in Figure 74:
Deliverable 3.2b
Page | 63 out of 107
Figure 74: T1 and T12 velocities at impact
Impact velocities are taken when complete thorax side (including shoulder) is in
contact with the buck. The overall trend that can be seen is that T1 and T12 x-
velocity and T1 z-velocity relative to the car increases when buck impact velocity
increases. The trend for T12 z-velocity is not so clear. The reason is that T12
rebound phase has not started, is just about to start or has already started.
Figure 75: Rib deflections on impact side.
Overall THUMS thorax deflection depends on buck geometry, impacting velocity and
impacting location but also if the arm on the impacting side is trapped between the
thorax and bonnet or not.
Thorax deflection on the impact side (right) is dependent on impact location, impact
velocity, buck geometry and THUMS kinematic. In all 30km/h load cases the left arm
get trapped between the thorax and the bonnet causing the higher rib deflections.
The highest deflection on the impact side is measured for the 30km/h MPV load case
Deliverable 3.2b
Page | 64 out of 107
Figure 76: Rib deflections on non-impact side.
On the non-impact side (left) the trend is an increasing deflection with increasing
impacting velocity and seems not to be effected by impact location. The highest
thorax deflection on non-impact side is measured for the 50km/h SUV load case.
3.2.1.2 TIPT As explained in Chapter 2.3.2, the torso of the ES-2 has been adopted as thorax
injury prediction tool (TIPT). To understand its viability as impactor, a broad set of FE
simulations has been performed. Several loops of simulations were done with the
aim to compare results from the most realistic HBM simulations, the TIPT simulations
best correlating to the HBM simulations and the most simplified TIPT simulations
under the same ambient conditions as the physical tests reported in Deliverable
D4.2(b). An overview of all TIPT simulations is given in Chapter 2.2, Table 2. Besides
the baseline simulations from loop 1, the most promising simulations in terms of
degree of correlation with HBM simulations (loop 3r1 and loop 5) are described in the
subsequent chapters. All remaining simulations are reported about in SENIORS
Deliverable D2.5b.
3.2.1.2.1 Loop 1 In order to compare the results between HBM and TIPT simulations, the impactor
was propelled at angles, speeds and arm positions identical to those of the HBM
thorax at the moment of impact.
A total of 12 tests were performed with the HBM, at four different vehicle speeds (20,
30, 40 and 50km/h) against three different models of the generic SAE Buck (Sedan,
Deliverable 3.2b
Page | 65 out of 107
SUV and Van/MPV). Therefore, also 12 TIPT simulations were performed, taking into
account the above mentioned variables.
To proceed with the TIPT simulations loop 1, the following actions were performed:
1. Design of the TIPT impactor model, cutting the Thorax of the ES-2 in LS-DYNA.
Figure 77: Reduced ES-2 dummy model and nodes for rib intrusions, rib accelerations and lower spine accelerations.
2. Identification of the key signals to be taken from the TIPT:
The output from the TIPT for comparison with the HBM results were rib intrusion,
spine acceleration and the tracking points. To match the HBM with the TIPT anatomy
for the tracking points, T1 and T12 were used:
Deliverable 3.2b
Page | 66 out of 107
3.
Figure 78: HBM and TIPT points used for tracking and speed matching.
Deliverable 3.2b
Page | 67 out of 107
Table 7: TIPT elements used for variables registration (tracking, acceleration and intrusion).
Item Component
Rib intrusions Change in length – Element 10500 Change in length – Element 10501 Change in length – Element 10502
Spine accelerations
Acceleration upper spine (T1) – Node 10264 (next to 10001 for global coord syst) Acceleration lower spine (T12) – Node 10461 (next to 10003 for global coord syst)
Tracking points T1 – Node 10001 (used ide history to obtain values in global coord. syst.) T11 (used T12 instead) – Node 10003 (used ide history to obtain values in global coord. syst.) T8, L5, T6 – are not possible to track using the TIPT
Figure 79: Elements used for rib intrusion registration.
Figure 80: TIPT elements used for tracking and speed registration.
Rib 1 – Element 10500
Rib 2 – Element 10501
Rib 3 – Element 10503
Lower spine Node 10461 for spine acceleration. Node 10003 for global coord syst.
Upper spine Node 10264 for spine acceleration Node 10001 for global coord syst.
Deliverable 3.2b
Page | 68 out of 107
3. Identify the thorax impact position:
The TIPT was rotated and translated so that the position of T1 and T12 were
approximately the same as for the HBM. It had been necessary to apply a move back
in order to avoid the intersection of the TIPT arm with the buck. Because of having a
big influence on the rib deflection, the arm was also positioned for each load case.
4. Calculate the thorax impact speed:
In the HBM simulations the SAE Buck impacted the stationary HBM, similar to a real
crash. The output from the accelerometers located at T1 and T12 of the HBM was
given in local coordinate system located in the corresponding positions. However, in
impactor simulations, the impactor was thrown against the bonnet, and therefore in
this case, the TIPT was moving against the stationary SAE Buck.
For simplification of the process, the output of equivalent nodes located near T1 and
T12 of the HBM were obtained in global coordinate system. Then the T1 and T12
speed were calculated in relative coordinates (over the vehicle). Finally, the average
of the relative T1 and T12 speed were calculated in order to apply the same launch
speed to all the TIPT nodes.
Deliverable 3.2b
Page | 69 out of 107
Initial velocity was applied to all the nodes
of the TIPT in x and z directions.
Vx and vz values were extracted from the
average relative velocity of T1 and T12 at
the instant of impact. The impactor moved
in free flight during the whole simulation.
Figure 81: T1 and T12 impact speed at absolute reference, relative reference and average as well as the TIPT impact speed description
5. Determine test configuration variables:
From the following table, summarizing the test configuration variables, the TIPT
impact parameters obtained from the HBM simulations were extracted.
Deliverable 3.2b
Page | 70 out of 107
Table 8: TIPT impact test characteristics.
SEDAN 20 km/h 30 km/h 40 km/h 50 km/h
Instant of impact (ms) 200 150 125 100 T1-T12 average relative translational velocity *with respect to the car **values in global coordinate system
20 km/h 30 km/h 40 km/h 50 km/h Instant of impact (ms) 165 110 90 75 T1-T12 average relative translational velocity *with respect to the car **values in global coordinate system
20 km/h 30 km/h 40 km/h 50 km/h Instant of impact (ms) 160 100 80 65 T1-T12 average relative translational velocity *with respect to the car **values in global coordinate system
interposed between bonnet and ribs resulting in smaller differences on peak rib
deflection, see Figure 94:
Figure 94: THUMS vs. TIPT time history ribs deflection for all tested load cases according to Table 11.
For the same reasons as described in Chapter 3.2.1.2.2 (rib extension phase in
THUMS and different TIPT kinematics during the impact), the time history ribs
deflection of THUMS and TIPT were hardly comparable with each other.
A correlation study using loop 5 simulations resulted in a reasonable correlation with
THUMS results, especially for the 4th rib, with a very good linear correlation:
Deliverable 3.2b
Page | 82 out of 107
Figure 95: THUMS vs. TIPT correlation on ribs peaks deflection
3.2.2 Actual vehicles
3.2.2.1 SUV After the TIPT model set up through impact tests on simplified vehicle models, the
simulation was carried out on an actual SUV representative.
Four tests were performed with TIPT impact points in different bonnet areas, using as
reference the test setup coming from loop 5 simulations. Three additional tests were
performed assuming as impact area the grille of the vehicle, this to replicate a
possible impact between vehicle with high front-end and e.g. child pedestrian.
The impact points on the SUV were identified through the WAD (Wrap Around
Distance) positions, as shown in Figure 96.
Deliverable 3.2b
Page | 83 out of 107
Figure 96: WAD distribution.
The tests configurations are summarised in Table 12:
Table 12: Summary of configurations for simulation.
WAD Y impact coordinate Impact speed Speed
angle Positioning
impact angle TEST 1 850 0 40 km/h 0° 0°
TEST 2 850 -SRL+133.5 40 km/h 0° 0°
TEST 3 850 +SRL-133.5, 40 km/h 0° 0°
TEST 4 1010 0 15 km/h 23° 70°
TEST 5 1190 0 15 km/h 23° 70°
TEST 6 1330 0 15 km/h 23° 70°
TEST 7 1330 +SRL -133.5 15 km/h 23° 70°
Deliverable 3.2b
Page | 84 out of 107
The TIPT impactor position has been identified as follows:
- The reference point for the alignment is defined by the intersection between
the middle rib plane and the vertical plane passing through the shoulder
reference point and the central rib (Figure 97).
- The TIPT was rotated and aligned on the WAD along the direction of the
velocity vector, as listed in the Table 12.
Figure 97: TIPT reference point
In Figure 98, an example of the alignment as used in Test 6 is depicted.
Figure 98: TIPT alignment in test 6.
Deliverable 3.2b
Page | 85 out of 107
The analysis of results comprised of the TIPT rib deflections and its lower and upper
spine accelerations.
Table 13: Ribs deflection and acceleration.
RIB DEFLECTION [mm] ACCELERATION [g]
UPPER RIB MIDDLE RIB LOWER RIB LOWER SPINE (T12)
UPPER SPINE (T1)
TEST 1 24.5 28.5 47 29.5 30.6
TEST 2 17.8 18.9 38.0 33.4 36.5
TEST 3 44.2 34.1 55.3 50.7 54.4
TEST 4 11 6.5 4 9.8 8.8
TEST 5 3.5 6.1 7.6 8.3 7.6
TEST 6 2.9 2.8 8.2 6.5 5.5
TEST 7 4.8 3.6 10 8.6 6.4
The results for each configuration are shown in the following images.
Figure 99: TEST 1 - WAD850, Y=0.
Deliverable 3.2b
Page | 86 out of 107
Figure 100: TEST 2 - WAD850, Y = –SRL+133.5.
Figure 101: TEST 3 - WAD850, Y = +SRL-133.5.
Deliverable 3.2b
Page | 87 out of 107
Figure 102: TEST 4 – WAD1010, Y = 0.
Figure 103: TEST 5 – WAD1190, Y = 0
Deliverable 3.2b
Page | 88 out of 107
Figure 104: TEST 6 – WAD1330, Y = 0.
Figure 105: TEST 7 – WAD1330, Y = +SRL-133.5
The results of the FE simulation were also compared with the experimental tests. A
description of the experimental tests is provided in SENIORS Deliverable 4.2(b).
Deliverable 3.2b
Page | 89 out of 107
In Table 14 the maximum values of the rib deflection (experimental and numerical)
are compared:
Table 14: Ribs deflection.
RIB DEFLECTION [mm]
UPPER RIB MIDDLE RIB LOWER RIB
EXP. RESULT NUM. RESULT EXP. RESULT NUM. RESULT EXP. RESULT NUM. RESULT
TEST 4 3.8 11 3.1 6.5 6.5 4
TEST 5 0.7(*) 3.5 2.8(*) 6.1 7.1(*) 7.6
TEST 6 0.4 2.9 2.1 2.8 6.2 8.2
TEST 7 2.0 4.8 3.7 3.6 6.6 10 (*) average value for the three repetitions
In Table 15 the maximum values of the spine acceleration (experimental and
numerical) are compared:
Table 15: Spine acceleration.
ACCELERATION [g]
LOWER SPINE (T12) UPPER SPINE (T1) EXP. RESULT NUM. RESULT EXP. RESULT NUM. RESULT
TEST 4 12.5 9.8 7.0 8.8
TEST 5 10.7(*) 8.3 10.4(*) 7.6
TEST 6 10.8 6.5 9.0 5.5
TEST 7 13.0 8.6 12.5 6.4 (*) average value for the three repetitions Both, simulations as well as experimental tests showed low levels of rib deflection
and spine acceleration.
Different considerations have to be made regarding the tests with a perpendicular
TIPT angle (test 1, 2 and 3). In these cases the different kinematics of the impactor
during the tests leads to high rib deflections and spine accelerations.
The readings of Test 2 were less severe compared to those of Test 3 due to the hood
profile near the SRL and the different area of interaction between ribs and frontend.
Deliverable 3.2b
Page | 90 out of 107
Figure 106: TEST 2 and 3
Deliverable 3.2b
Page | 91 out of 107
4 SUMMARY AND CONCLUSIONS
4.1 CONCLUSIONS OF VALIDATIONS AND IMPACT ON SENIORS PROJECT This report resumes the results of baseline pedestrian simulations with human body
models and impactor models against generic test rigs. Subsequent to the work
reported about in Deliverable D2.5b, a number of impactor simulations vs. four actual
vehicle models and a generic SAE Buck and its derivatives (representing a Sedan,
an SUV and a Van/MPV frontend) have been carried out and compared to the results
from THUMSv4 human body model simulations against identical frontends. The
Deliverable compares kinematics, time histories as well as peak loadings and
identifies possible correlations between the loadings to the human body model and
the impactor models.
The simulations reported about in this Deliverable and related to the assessment of
lower extremity injuries state an altogether good quantitative correlation in terms of
time histories as well as impact kinematics between the THUMSv4 human body
model and the FlexPLI derivatives with applied upper body mass (FlexPLI-UBMrigid
and FlexPLI-UBMrubber), with some further advantages of the FlexPLI-UBMrubber, as
shown in chapter 3.1.3.1 (compare Figure 50), and significant advantages over the
FlexPLI Baseline.
However, the good qualitative correlation cannot always be demonstrated with the
quantitative correlation of peak femur and tibia bending moments and maximum knee
ligament elongations. The obviously good quantitative correlation between the
FlexPLI Baseline and HBM femur bending moments is not in line with the
obeservations made during the development of the FlexPLI. The sometimes poor
quantitative correlation between FlexPLI-UBMrigid / FlexPLI-UBMrubber and HBM, in
particular in terms of femur bending moments, can be partly explained by premature
or unintended interaction between the upper body surrogate and the vehicle’s front
leading edge, resulting in incorrect timings and magnitudes of the peaks. Removing
those results contributes to significant better quantitative correlations, as also
demonstrated in Figure 62 and Figure 63.
Deliverable 3.2b
Page | 92 out of 107
Application of an upper body mass to the FlexPLI contributes to an improved
kinematics not only at vehicle centreline, but also towards the ends of the bumper
beam.
Finally, the FlexPLI with UBM shows a superior behaviour over the FlexPLI Baseline
regarding time histories and impact kinematics. When avoiding premature or
unintended loading of the mass, also the peak values correlate well with the HBM in
most cases.
Remaining action item is the establishment of impactor limits to be derived from
human injury risks as described in Deliverable D2.5b, for finally using the FlexPLI-
UBMrubber within an assessment procedure that will be reported about in Deliverable
D4.1(b). In this context, also the appropriateness of peak values as injury indicators
needs to be further discussed.
The simulations reported about in this Deliverable and related to the assessment of
thoracic injuries show the in principle good approach of using the isolated ribcage of
the ES-2 dummy during component tests. However, several limitations need to be
taken into account at this point in time. First of all, the ES-2 dummy was designed as
vehicle occupant for the assessment of lateral impacts. HBM simulations with TUC
THUMS show the kinematics of the human thorax differing from this loadcase, with
oblique thorax angle and the velocity vector not perpendicular to the thorax. The
capacity of capturing oblique loadings with the available instrumentation is limited.
THUMS rib extension before the impact, rotational elements in THUMS kinematics
and translational movement of the TIPT furthermore result in different impact
locations and loadings. Thus, besides the diverging time histories of THUMS and
TIPT rib intrusions, also the quantitative correlations are unsatisfactory in most
cases. When using a TIPT based on the ES-2 ribcage, establishing impactor limits
should rather be based on injury criteria for the ES-2.
Deliverable 3.2b
Page | 93 out of 107
4.2 TRANSFER OF RESULTS The FE simulations reported in this Deliverable will be used as input parameters for
the definition of draft test and assessment procedures as described in Deliverable
D4.1(b). Furthermore, the setups will be replicated during physical tests reported in
Deliverable D4.2(b). Regarding the FlexPLI and its derivatives with upper body mass,
the test setup is, in principle, well established and described in Euro NCAP (2017)
and UNECE (2015). Furthermore, correlations between FE simulations and testing
will be investigated and the draft procedures will be finalized, based on further
insights during the physical tests.
The TIPT physical tests will be performed under the same conditions as the last
simulation loop. This section describes the transfer of the variables and results from
the simulation to the physical tests reported in D4.2(b).
1. Test tool and acquisition values:
The test tool consists of the thorax/ribcage of the ES-2 with only one arm (RHS) and
no abdomen. The arm is parallel and stowed to the body. Moreover, a support to be
attached to the launcher bracket is to be added. The total mass for this configuration
is 22.5 kg. The TIPT should be instrumented with the IR-TRACCS in order to obtain
the rib deflections and with two accelerometers close to the theoretical T1 and T12
locations in order to compare the impact speed values.
ES-2 thorax
simulation model TIPT model with
one arm TIPT impactor drawing with
supports and bracket
Figure 107: TIPT model evolution.
Deliverable 3.2b
Page | 94 out of 107
2. Impact speed:
According to the impact parameters as defined in Chapter 3.2.1.2.1 par. 5, the impact
speed of the TIPT and angle of the velocity vector at vehicle speeds of 40 km/h are