Final Report Summary - SAFEEV (Safe Small Electric Vehicles through Advanced Simulation Methodologies)
Executive Summary:
The project SafeEV, “Safe Small Electric Vehicles through Advanced Simulation Methodologies”, was funded by the European Community’s Seventh Framework Programme under the Transport programme (Grant Agreement no: 314265). It consists of 11 partners from 5 different countries. Our consortium combines industry (four OEMs, one TIER 1 supplier, one SME) and research (five research organisations) in the field of vehicle safety.
The number of small and light-weight full electric vehicles will substantially increase especially in urban areas. These Small Electric Vehicles (SEVs, in SafeEV in particular concern vehicles in the mass category of so-called L7e vehicles) show distinctive design differences compared to the traditional car (e.g. no bonnets, vertical windscreens, outstanding wheels). Thus the consequences of impacts of SEVs with vulnerable road users (VRU) and other (heavier) vehicles will be different from traditional collisions. These fundamental changes are not adequately addressed by current vehicle safety evaluation methods and regulations. VRU protection, compatibility with heavier opponents and the introduction of active safety systems have to be appropriately taken into account in order to avoid any SEV over-engineering (e.g. heavy or complex vehicle body) by applying current regulations and substantially impair the SEVs (environmental) efficiency.
Based on future accident scenarios, the project SafeEV aims to define and develop advanced test configurations and evaluation criteria for VRU protection, occupant safety and compatibility of SEVs. Moreover, industrial applicable methods for virtual testing of these configurations and criteria (e.g. a method for active occupant safety assessment) will be developed. These methods are applied in order to derive protection systems for enhanced VRU and occupant safety for SEVs. The evaluation of one developed hardware system was used to demonstrate the potential and applicability of these new methods. Dedicated best practice guidelines for VRU and occupant safety evaluation of SEVs will ensure a sustainable impact for industry and regulatory organisations beyond the project duration. With the new evaluation methods developed, vehicle safety for SEV on urban roads in the next decade will be adequately addressed resulting in less fatalities and injuries without compromising vehicle efficiency. Moreover cost-efficient development of SEVs will be made possible by the new virtual testing methodologies developed.
Project Context and Objectives:
WP 1:
The objective of the work package is the identification of the most relevant future accident scenarios involving small electric vehicles (SEVs). These SEVs are coming to the road in the near future and therefore prediction of the most relevant accident scenarios of these vehicles has to be made. SafeEV has focused on the crashworthiness of SEVs looking into mitigation of the crash-severity during pre-crash and crash. A focus is set on the influence of pre-crash safety systems on future (next decade) accident scenario’s as well as on pedestrian (vulnerable road user) protection and compatibility for future small electric vehicles. (pre-crash and compatibility are addressed in Matisse project and transferred to SafeEV).
Two Tasks were defined within WP1:
• Task 1.1 Methodical analysis on future pedestrian accident scenarios involving SEVs (M1-6)
• Task 1.2 Conclusions on future accident scenarios including results from MATISSE (M7-8) [Input from MATISSE]
Task 1.1 Methodical analysis on future pedestrian accident scenarios involving SEVs (M1-6)
Summarized the main assessment steps for the prediction on future accident scenarios is summarized with the following bullet points.
• Accident Type Assessment
• Efficiency Analysis Method for Driver Assistance Systems
• Stochastic Accident Prediction Approach
• Delphi Study (Expert Survey)
• Public Survey
• Pedestrian Kinematics Analysis including New Car Geometries
Task 1.2 Conclusions on future accident scenarios including results from MATISSE (M7-8) [Input from MATISSE]
Task 1.2 was to summarize the most important findings of the Matisse D1.1 and the SafeEV D1.1 documents.
Matisse D1.1 was focused on occupant related scenario prediction. The study also included a review of current Swedish car accidents and their relevance for future scenarios in view of forecasted safety systems developments. In addition accident data from the German In-Depth Accident Study (GIDAS) database were extracted. Smaller M1 vehicle models were categorised as Small Urban Vehicles and compared to a matched set of heavier M1 vehicles. Only accidents that occurred in conditions that would be likely for a small urban vehicle were included (maximum 2 occupants, maximum speed 100 km/h, maximum travel distance 100 km).
WP 2:
WP 2 “Advanced test configurations and criteria for pedestrian and occupant safety in SEVs” comprised three tasks:
Task 2.1: Definition of test conditions for vulnerable road user protection in crashes involving small electric vehicles
Task 2.2: Definition of test conditions for occupant protection (incl. compatibility) assessment in crashes involving small electric vehicles
Task 2.3: Criteria for assessment of SEV safety in crashes (M9-12)
The objectives of WP2 are:
• Development of necessary test configurations (accident conditions) for vulnerable road user and occupant protection assessment (incl. compatibility) in accidents involving SEV’s.
• Specification of the methodologies and tools for safety assessment in the proposed test set-ups (experimental, virtual as well as hybrid approaches).
• Assessment of the applicability of current evaluation criteria for SEVs and promote the development of new criteria where current criteria are inadequate to address SEVs safety. This concerns both vehicle based criteria incl. fire and electrical safety criteria and injury assessments criteria. The criteria should be suitable for virtual testing.
The test conditions are based on the results of WP1 of SafeEV, recent regulatory trends and new developments in the field of consumer testing (NCAP). The assessment of pre-crash based injury reduction systems is also taken into account. The work included specification of the (virtual) tools to be used like impactors, virtual models, crash dummies, integrated experimental/virtual methods, as well as criteria to be used for the assessment. A first proposal of test scenarios for frontal and side impacts (occupant protection) as well as a proposal for pedestrian and cyclist protection for SEVs in urban environment were developed.
WP 3:
Work package 3, Advanced simulation methodology for integrated pedestrian / occupant safety in SEVs, overall goal is to develop a seamless tool chain to investigate solutions for pedestrian safety and occupant protection towards virtual certification testing. Therefore the following objectives will be pursued:
• Develop REVMs for the development of the advanced evaluation methodologies
• Develop simulation and testing tools for pre-crash and crash safety evaluation needed for the test configurations and criteria from WP2
• Develop tool chain for pedestrian safety analyses (based on existing human models which will be enhanced within this project by use of brain model, active safety evaluation, ...)
• Develop tool chain for occupant protection incl. compatibility analyses (using e.g. generic vehicle fleet on the test configurations and criteria from WP 2)
• Develop pre-crash sensing models and integrate them into virtual vehicle models of SEVs in order to create virtual pre-crash-data and evaluate feasibility of optimized sensing technologies and topologies.
• Final revision of the test configurations and criteria from WP2
The work package is splitted into four tasks
Task 3.1: Setup of two reference electric vehicle models (REVM)
Task 3.2: Development of an advanced simulation methodology for consistent safety analysis for pedestrian protection for SEVs
Task 3.3: Development of an advanced simulation methodology for a consistent safety analysis for occupant protection (incl. compatibility) in SEVs
Task 3.4: Revision of test configurations and evaluation criteria towards the new developed evaluation methods
This work package was started in project month 4 (January 2012). The first task was the setup of two reference electric vehicle models (REVM). These two models are essential for the whole simulation work, because these are used for further activities for the development of the advanced evaluation methods. Two different models were generated: one, the REVM1, was conceived since the beginning as a native electric car while the other, the REVM2, was derived from an existing ICE small vehicle. Both REVMs were originally built within LS Dyna F.E. code; subsequently, only the REVM2 was translated into PAM-Crash code, too. Task 3.2 and 3.3 were focusing in the development of an advanced simulation methodology for pedestrian and occupant safety and define a virtual tool chain, results are summarized in deliverable D3.5 and D3.6. Within D3.7 the revision of test configurations and evaluation criteria towards the new developed evaluation methods, based on D2 and updated with findings of WP3.
WP 4:
Work package 4 is dedicated to the definition of generic safety solutions to improve the safety level of SEVs with respect to occupant and pedestrian protection and the assessment of the defined measures applying SafeEV’s virtual evaluation tool chain. Two subtasks are defined for work package 4:
Task 4.1: Definition and assessment of advanced pedestrian safety measures for SEVs.
Task 4.2: Definition and assessment of advanced occupant safety measures for SEVs.
The objectives of the two tasks within work package 4 are as defined:
• Definition of generic improvement measures to augment the safety level of SEVs with respect to occupant and pedestrian protection
• Application of the virtual evaluation tool chain defined within SafeEV project work package 3 to investigate the effect of the respective measures on the fulfilment of the crashworthiness target criteria defined for SEVs as represented by the REVMs used within the project
• Identification and more detailed elaboration of the most promising solutions focusing on the crashworthiness level and lightweight targets for SEVs
• Formulation of guidelines for occupant and pedestrian protection for future SEV design activities summarising the lessons learnt during the execution of the different study steps within work package 4
All planned objectives have been fulfilled.
The results of task 4.1 are documented within SafeEV project deliverable no. D4.1 “Generic safety solutions and design candidates evaluated with the advanced methodology for pedestrian safety”. The reporting of the results of task 4.2 can be found in SafeEV project deliverable no. D4.2 “Generic safety solutions and design candidates for occupant protection including compatibility improvements”. In general it can be stated that various generic safety solution were performed and the most promising component were identified. This component is then be evaluated within WP 5.
WP 5:
The workpackage is titled, “Use case for enhanced safety for SEVs using virtual testing” and is split into two subtasks:
Task 5.1 Selection of the use case and dedicated design (M31-34)
Task 5.2 Evaluation of the use case (M34-35)
This work package elaborates a use case for an advanced safety solution and applies the evaluation methodology developed in this project. In T5.1 a pedestrian protection solution investigated in WP4, a stiffness modifiable transversal beam, has been selected for further development and physical prototyping. The manufactured beams have been produced and small scale versions tested as well as the originally beam. Furthermore, two more REVM1 concept solutions are built up to the level of virtual demonstrators and the front structure of REVM1 has been analysed with respect to signals. In T5.2 the virtual version of this beam has been included in the FE model and used to perform the virtual testing.
WP 6:
The main objective of this work package was to promote dissemination and exploitation of the project findings and results through organisation of dedicated workshops and information exchange activities and presentation at relevant conferences and scientific publications. One main task is formulated in Annex 1 of the project for WP6 which addresses dissemination and exploitation activities and the role of the partners in this area. Most of the activities started in month one. Five categories have been processed:
• Establishing project oriented dissemination & information platform (website, leaflet, etc). (in agreement and coordination with WP8 / SEAM activities)
• Contribution to and publication via conferences and scientific journals
• Implementation and exploitation of the main project findings to interest groups and relevant stakeholders outside the consortium.
• Deliverables (public) – as reference for further & subsequent activities
• Dedicated workshops (Final dissemination workshop)
WP 8:
The main objectives are communication within the project (meetings, reviews, decision making process, technical project management), the coordinate activities within the SEAM cluster and to provide the best practice guideline as one of the major project outcomes.
The work package is substitute into 3 tasks. In task 8.1 all project results will be reviewed and consolidated aiming at the formulation of a best practice guideline for the implementation of the advanced simulation methodologies into and “tool chain” for enhanced safety analysis concerning pedestrian and occupant safety in SEVs.
Task 8.2 comprises all activities within the SEAM cluster (SEAM project office, liaison team and dissemination strategy, starts in month one and ends in month 36. The SEAM cluster was initiated in order to coordinate and harmonise the four projects SafeEV, ENLIGHT, ALIVE and MATISSE. Main purpose of the SEAM cluster is to realise and monitor synergies between the four projects on RTD and demonstration level and to execute joint dissemination and exploitation activities.
Task 8.3 is the technical project management and communication and infrastructure for quality monitoring, decision making and conflict resolution.
Project Results:
D1.1:
Within Task 1.1 of WP1 most relevant future pedestrian accident scenarios involving small electric vehicles had to be estimated. This directly addresses next decade small electric vehicles (SEV´s) which will be seen in urban areas. The methodology used to identify the scenarios includes four main steps. Analysing national and in-depth accident databases give an overall scenario definition and the boundaries for a generic accident type. Stochastic simulations of the defined scenario show a trend of future impact speed. A Delphi study, supported by a public survey, estimates future mobility and city-layouts and helps to identify boundary conditions. And finally the fourth topic pedestrian kinematic analysis to evaluate the effect of future vehicle shapes and precrash systems.
From the database analysis six accident scenarios where seen as most relevant for future urban traffic, which includes near- and farside accidents before/after and outside of junctions (vehicle going straight). A generic simulation model is built up for a stochastic analysis, where the results show an overall decrease of speed with a pedestrian protection system on-board. The results of the Delphi study confirmed the assumptions made for future traffic scenarios. The consulted experts predict a decrease of average vehicle weight with a higher share of SEV´s and a dynamic reduction of future speed limits. Furthermore SEV´s being produced after 2017 will be equipped with enhanced safety features which will provide better pedestrian protection. Public transport will be improved and will have higher shares. The demographic change and the trend of aging in cities will also influence accident scenarios, but should be partly solved with an improved traffic management. The public survey shows similar trends with only a few deviations (e.g. estimated higher share of driver assistance systems). In the pedestrian kinematics analysis four future SEV designs are compared with conventional passenger cars with the focus on head contact including impact and post impact phase. It is shown that the impact speed of an adult and child head reduces with a high share of windscreen contacts due to vehicle shapes. Thus, a reduction of impact speed for future pedestrian accidents is expected. Most frequent predicted velocities will be between 10 km/h and 30 km/h.
D1.2:
D1.2 summarizes the most important findings of the Matisse D1.1 and the SafeEV D1.1 documents. The two projects are part of the SEAM cluster and they collaborate on the analysis and prognosis of future accident scenarios involving small alternatively powered vehicles (APV) in urban areas.
The methodology of the Matisse D1.1 and the SafeEV D1.1 included the following steps: 1) Analysing national and in-depth accident databases give an overall scenario definition and the boundaries for a generic accident type. 2) Stochastic simulations of the defined scenario show a trend of future impact speed with a generic pre-crash system on-board. 3) A Delphi study, supported by a public survey, estimates future mobility and city-layouts and helps to identify boundary conditions. 4) Review of current heavy truck accidents and their relevance for future scenarios in view of forecasted developments of safety systems. 5) Identification of critical accidents for high-pressure storage tanks by expert brainstorming and finite element methods. 6) Pedestrian kinematic analysis to evaluate the effect of future vehicle shapes and pre-crash systems. Four future SEV designs were compared with conventional passenger cars with the focus on head contact including impact and post impact phase.
The present report also included a review of current Swedish car accidents and their relevance for future scenarios in view of forecasted safety systems developments. In addition accident data from the German In-Depth Accident Study (GIDAS) database were extracted. Smaller M1 vehicle models were categorised as Small Urban Vehicles and compared to a matched set of heavier M1 vehicles. Only accidents that occurred in conditions that would be likely for a small urban vehicle were included (maximum 2 occupants, maximum speed 100 km/h, maximum travel distance 100 km).
Conclusions, Car to Car Accident Scenarios:
The stochastic analysis results show an overall decrease of the number of crashes and a decrease of the collision speed. For future cars of lower mass, crash delta-v will however increase. The results of the Delphi study predict a decrease of average vehicle mass with a higher share of APV´s and a dynamic reduction of future speed limits. The public survey shows similar trends with only a few deviations (e.g. estimated higher share of driver assistance systems).
Conclusions, Car to Pedestrian Accident Scenarios:
Six pedestrian-to-car accident scenarios where seen as most relevant for future urban traffic, which includes near- and far side accidents before/after and outside of junctions (vehicle going straight). A generic simulation model using stochastic analysis show an overall decrease of collision speed with a pedestrian protection system on-board. The results of the Delphi study confirmed the assumptions made for future traffic scenarios. A higher share of SEV´s and a dynamic reduction of future speed limits was forecasted. Furthermore APV´s being produced after 2017 will be equipped with enhanced safety features which will provide better pedestrian protection. It is shown that the impact speed of an adult and child head reduces with a high share of windscreen contacts due to vehicle shapes. Thus, a reduction of impact speed for future pedestrian accidents is expected. Most frequent collision speeds are forecasted to be in the range 10 km/h to 30 km/h.
In the present work several different and independent data sets and methodologies have been applied. They all include significant uncertainties. It is thus reassuring to be able to conclude that they all forecast similar trends. This gives the results a significant robustness.
In summary, it becomes obvious that certain collision types will be reduced in numbers and in average collision severity, but in all cases an estimated residual of likely 10% or more will remain and there will be a significant remainder also of severe collisions. The general publichave become used to the current vehicle safety level and will have very limited acceptance for increased injury risks as a result of the downsizing of the car fleet. If people lose trust in the new generations of small light vehicles they will refuse to buy them and they will penalise politicians that try to enforce their introduction. It must therefore be concluded that SEVs or APVs developed for production in 2025 must meet at least the same occupant and pedestrian crash protection requirements as for current vehicles
D2:
The objective of D2 is the specification of test configurations for vulnerable road user and occupant protection assessment (incl. compatibility) in accidents involving small EV’s in urban areas. Small EV’s in SafeEV in particular concern vehicles in the mass category of so-called L7e vehicles, where currently hardly any safety requirements exist.
The test conditions are based on the results of WP1 of SafeEV, recent regulatory trends and new developments in the field of consumer testing (NCAP). The assessment of pre-crash based injury reduction systems is also taken into account. The work includes specification of the (virtual) tools to be used like impactors, virtual models, crash dummies, integrated experimental/virtual methods etc...
Concerning evaluation of pre-crash and active safety systems, discussed in Chapter 2, it was concluded that the future focus in SafeEV will be on passive safety pedestrian and crash sensors and the according requirements for sensor evaluation tests. For pedestrian safety sensors it was shown that through their geometry SEVs pose a challenge on the setup of the sensors as well as on the sensor and trigger time of in-crash safety systems for pedestrians. Further it was concluded that only simulation/virtual tools will allow a continuous observation and evaluation of integrated safety systems in terms of a benefit based assessment.
For VRU (i.e. pedestrian and cyclist) safety evaluation in SafeEV, presented in Chapter 3, simulations are proposed using human body models (HBM) in the following 4 sizes: 6 year old child, 5% Female, 50% Male and 95% Male. Simulations concern impacts of a pedestrian against the vehicle front at 2 speed ranges: 25 km/h to 30 km/h (lower boundary) and 45 km/h to 50 km/h (upper boundary). Also simulations for cyclist protection are proposed.
Occupant protection is discussed in Chapter 4. The following accident types have been addressed: frontal, side, rollover, rear, compatibility and Multiple Impact Crashes (MIC’s).
For frontal occupant protection for L7e class vehicles a test (simulation with HBM) with an oblique impact configuration (30 degrees) is proposed. One option here is that also accounts for compatibility aspects, is a test with at a test speed of 30-40 km/h against a movable barrier that represents the average opponent vehicle mass (1125 kg in GIDAS analysis). Such a configuration also adapts automatically the effect of variations in SEV mass. The option for a full virtual test method here has high potential due to the fact that for crash barriers already validated models are available. Verification and Validation issues concerning the vehicle model are in general addressed by the IMVITER project and should be taken into account.
Concerning future frontal compatibility protection of M1 vehicles the ODB offset test configuration proposed by the FIMCAR project as well the FWDB test, are identified as most suitable tests to be used within the SafeEV project. The quantification of the use potential of a test configuration using a movable barrier for compatibility tests addressing SEV-to-SEV and SEV-to-M1 vehicle front crash behaviour will be an aspect addressed within the project work package 3.
For side impact occupant protection in L7e vehicles a test (simulation with HBM) using the latest MDB with a mass of 1100kg reflecting typical future car masses, is proposed. The barrier will hit the vehicle with a speed of 40 km/h at an angle of 90º. The impact location is moved backwards reflecting a short bonnet in a car body.
Rollover protection is not an important accident scenario in Europe (compared to the USA). Within SafeEV it is proposed that the SSF (Static Stability factor) and AST (Airbag standing Time) are calculated. For this no simulations or tests are needed. Eventually if these values exceed certain critical values it could be considered to perform so-called Fishhook and ESC tests which can be done by simulations. An Inverted Roof Crush test could be considered as well but will not be addressed within SafeEV.
Concerning rear impacts the introduction of AEB and introduction of seat concepts aimed at lowering the risk of WAD (Whiplash Associated Disorders) will reduce the number and risk of WAD in future rear-end impacts. However WAD will remain frequent and the consequences will continue to lead to large societal costs and personal suffering. Additional and improved test tools and sled test conditions are under development. Among these are a female dummy and associated limits and an adult 95Th percentile dummy. Also there is potential for simulations with HBM’s. Initial positioning of the occupant is a parameter having a large influence.
A final accident condition that was considered are the Multiple Impact Crashes (MIC’s), in which a vehicle experiences at least 2 impacts after each other. Although the frequency and injury severity of MIC’s will decrease due to the trends presented in WP1, a strong need to protect people in MIC’s remains due to the large frequency of these crashes and the relative high injury risk.
Recommendations concerning future work in SaveEV (WP3 etc...) concerning MIC’s are: (1) Inclusion of occupant simulations (virtual testing) with active HBM of extreme controlled vehicles manoeuvres that are aimed to reduce the risk of or the severity of second (or more) impacts in a MIC. The aim of such simulations is to study the effect on human position changes during such manoeuvres and the resulting risk on OOP’s and (2) Inclusion of occupant simulations aiming of optimization of reversible and irreversible restraint systems during MIC’s.
Chapter 5 deals with criteria to be used in the various test (simulation) conditions. A distinction is made between injury criteria, compatibility criteria and fire and electric safety criteria.
The overview of injury criteria is rather extensive. Most important body parts and future trends are addressed (including the need to have better criteria for children and elderly). Both global criteria (criteria that also can be determined at physical crash dummies) and criteria on tissue level to be determined on human body models are described. Also a distinction is made between criteria for vulnerable road users (pedestrians and cyclists) and car occupants.
The compatibility evaluation criteria defined within the FIMCAR project for the FWDB and the ODB test are initially identified as also suitable for the assessment of small electric vehicles. But it should be realized that FIMCAR didn’t integrate a force compatibility assessment criterion into the final metric proposal, so this may need some additional attention.
Concerning fire and electric safety criteria an overview of current regulations for crash safety of Battery Electric Vehicles (BEV) is given. All requirements are for an experimental evaluation procedure, focusing on frontal, side and rear impacts. A virtual assessment of these criteria is not considered yet. Possible reasons might be the limited knowledge of or the experience with the crash safety of BEV and the lack of suitable simulation models. A review of trends in this field did not predict big changes in the current requirements for 2025.
D3.1:
D3.1 describes the Reference Electric Vehicle Models (REVMs) with respect to their requirements, capabilities and validation status.
These REVMs are used within the Project as generic analysis tools for the development of those advanced evaluation methodologies that will form the pillars of a virtual tool chain to be used in future virtual certification procedures/protocols specifically dedicated to small electric vehicles, an aspect that represent one of the main goals of SafeEV.
Such models are intended to represent possible small electric vehicles including solutions that will be distinctive for the next generation of SEVs, i.e. new designs which are quite different from current traditional cars. New concepts of Small Electric Vehicles for urban areas (short SEVs) already presented to the public might belong to the M1 category (motor vehicle for passenger transport with not more than eight seats excluding driver’s seat) or to the L7e category (four-wheeled vehicle up to 400 or 500kg). Then, the approach followed in REVMs generation was to model very small M1 category vehicles that can be used even as representative of L7e category, through an adequate and artificial reduction of the total vehicle mass. In fact, this is a possible and practical option especially for those numerical simulations where pedestrian impact configurations are involved.
Two different models were generated: one, the REVM1, was conceived since the beginning as a native electric car while the other, the REVM2, was derived from an existing ICE small vehicle. Both REVMs were originally built within LS Dyna F.E. code; subsequently, only the REVM2 was translated into PAM-Crash code, too, due to specific operational needs of T3.2 related to pedestrian impact simulations with human body models. The details about such model translation are described within deliverable D3.4.
These models can be used to evaluate the behaviour of the crash structure (e.g. crash pulse, deformation characteristics and intrusions) in typical car-to-barrier and car-to-car impacts.
Moreover, they can be used for the simulation of pedestrian impacts (through pedestrian impactors and standing pedestrian dummies and/or Human Body Models).
However, no restraint systems are included in these models thus no assessment of occupant readings is possible with these first releases of REVMs: such (generic) restraint systems will be added in a subsequent phase of the project (T3.3) in order to perform the integrated biomechanical/structural work required when studying compatibility aspects of SEVs.
In any case, for the preliminary assessment of the occupant loading conditions with the first REVM releases described in this deliverable, the evaluation of the crash pulse obtained from their numerical crash simulations remains possible, by the application of known severity indicator like the OLC (Occupant Load Criterion).
The simulation runs performed so far showed that the REVM models are numerically stable and behave in a realistic way.
D3.2:
WP3 of SafeEV deals with an advanced simulation methodology for integrated pedestrian / occupant safety in small electric vehicles (SEVs). D3.2 is part of Task 3.2 of work package 3 and concerns the development of an advanced simulation methodology for consistent safety analysis for pedestrian protection for SEVs. Task 3.2 runs in parallel with Task 3.3 which deals with the corresponding development of a simulation methodology, but for occupant protection only. The objective of the D3.2 report is as an initial definition of requirements for consistent safety analysis for pedestrian protection in SEVs. These requirements are interpreted as the necessary steps to develop a so called “seamless tool chain” in order to virtually assess and optimise pedestrian safety. The main aspects covered by this report are the following:
• test conditions and body regions to be evaluated (Chapter 2)
• simulation tools to be used by the different partners (Chapter 3)
• a brief look back to the EC funded project IMVITER (IMplementation of VIrtual TEsting in safety Regulations) in order to provide an exemplary process and protocols which could enable type approval through virtual testing (Chapter 4)
• description of the necessary steps and tools needed to develop a “seamless tool chain” for pedestrian safety (Chapter 5) consisting of:
- agreement on how to evaluate injury risk using human body models (HBMs)
- comparison of a HBM vs. a pedestrian accident compliant (PAC)
- code dependence in the application of Finite Element (FE) HBMs
- utilisation of an advanced ground model to evaluate the secondary impact
- comparison of acceleration sensor signals derived from pedestrian impactor and HBM contact against the vehicle
The reporting of Task 3.2 will be completed by three future reports. Report D3.4 will make a comparison of the simulations with the models running under different FE codes for identical load cases. Thereafter, report D3.5 will describe the resulting methodology for the virtual tool chain for pedestrian safety simulation. This will include an analysis of the robustness of the virtually gained results by slightly varying the defined load cases. Finally report D3.7 will summarise the final definition of the relevant load cases and appropriate criteria for injury risk evaluation using HBMs as a result of a separate Task 3.4 in WP3. Hence report D3.7 finally closes Tasks 3.2 to 3.4 and WP3 itself.
D3.3:
The work in WP3 concerns the development and evaluation of virtual test methods and simulation tools for small electric vehicles. D3.3 is the first deliverable of task 3.3. The general objective of this report is to define the requirements for the development of an advanced simulation methodology for a consistent safety analysis for occupant protection (incl. compatibility) in SEVs.
The aim of this report is to specify:
• the test conditions to be evaluated in task 3.3 (Chapter 2);
• the simulation tools to be used in task 3.3 (Chapter 3);
• the characteristics that a protocol for type approval through virtual testing shall have, according to the work developed by the IMVITER (IMplementation of VIrtual TEsting in safety Regulations) project (Chapter 4); and
• the description of tools that will be developed in the remaining part of task 3.3 to complete the virtual tool chain (Chapter 5).
The report includes the simulation test matrices which states which accident scenarios will be evaluated and by whom. Other important information is the specification of which criteria to use in the evaluation. The matrices and criteria are specified in Chapter 2. The criteria are not finally concluded, but an important activity of task 3.3 is to further develop and agree on relevant criteria. Based on these criteria, needs to develop suitable post-processing tools will be confirmed and/or arise. All these mentioned activities will be worked on jointly and in parallel by the partners. A description of who does what in general terms is provided in Chapter 5. The final results of task 3.3 will be included in the D3.6 report of SafeEV.
D3.4:
WP3 of SafeEV addresses the development and demonstration of a seamless tool chain in order to virtually develop, optimize and assess integrated occupant and pedestrian safety in small electric vehicles. Within Task 3.2 the main focus is on the development of an advanced simulation methodology for consistent safety analysis for pedestrian protection and the implementation of WP2 scenarios and evaluation criteria. FE- Human Body Models (HBM) are identified as possible candidates for such an application and already described in detail in deliverable D3.2 (pedestrian). Finally the HBM THUMS and also the REVM2 (see D3.1) are used by two project partners in similar test scenarios applying similar criteria – but utilizing different FE solver, namely LS-Dyna and PAM-CRASH/VPS.
In terms of a reliable and universal tool chain it is essential that these simulations and models generated under different codes deliver comparable results concerning a dedicated criterion – in this case injury risk prediction.
D3.4 will therefore explicitly discuss the code dependency, translation and finally harmonisation aspects in the application of FE-HBM and the projected REVM.
In the first part of the report simulation results, generated under identical load cases, of THUMS-D (Ls-Dyna) and THUMS-VW (VPS) are compared. Following recommendations from AMSE concerning V&V (Verification & Validation) this was exemplarily done on different discretisation levels of the HBM: Component level (here: Tibia & Femur) > body region level > full body level / kinematics. Also the implementation of SUFEHM (Strasbourg Univ. Finite Element Head Model - incl. post-processing tool) under different codes and the development of code specific criteria are discussed within this chapter.
The second part of the report addresses the code dependency and translation aspect of the vehicle model REVM2. Results from identical load cases (impactor tests) were compared for REVM2 running under LS-Dyna and the REVM2 under VPS.
The comparison shows, that especially the vehicle model in the two codes delivers quite comparable results when standard impactors are used.
Even though the THUMS models were not “similar” in this study, it could be stated, that the general injury risk prediction given by the two models is quite comparable. SUFEHM, as “identical” body part, has shown similar risk prediction for two injury mechanisms. Nevertheless, and this might be the reason for the discrepancy concerning the third injury mechanism, the head contact is also influenced by slightly different full body kinematics.
Further on a comparison of simulation results in terms of comparable prediction of injury risk will also be done with the PAC model and the REVM 1 w.r.t THUMS-D in the same impact condition. This will be reported in D3.5.
D3.5:
Work package WP3 deals with the “Advanced simulation methodology for integrated pedestrian / occupant safety in Small Electric Vehicles (SEVs)”. The work package is split into four tasks and mainly describes the development of the virtual tool chain for occupant and pedestrian safety evaluation.
D3.5 is a follow-up on the D3.2 report which defined the requirements of the models in the virtual tool chain to be applied to the test setups for the pedestrian protection evaluation as defined in report D2. Hence, the report D3.5 completes the documentation of the work carried out under task T3.2 “Development of an advanced simulation methodology for consistent safety analysis for pedestrian protection for SEVs”.
The main objective of this deliverable D3.5 is to report on the application of the selected models within the virtual tool chain for pedestrian protection. First simulations of the relevant load cases have shown that further specifications of boundary conditions for the simulations became necessary. These are documented in the introductory chapter of this report along with descriptions of the updates as for the vehicle models, the Human Body Models (HBMs) and the Strasbourg University Finite Element Head Model (SUFEHM). The head injury risk assessment based on UNISTRA’s SUFEHM was enhanced during the course of task T3.2 by the extension of implied real world head trauma data together with an update of the Injury Risk Assessment (IRA)-tool.
In the following section the project partners Daimler and VW report on the application of HBMs as there are THUMS-D (Daimler) and THUMS-VW respectively which represent the main part of the virtual tool chain for pedestrian safety analysis. In some cases identical impact conditions were simulated and analysed such that a comparison of the results obtained by using similar HBMs under different crash codes was possible as well. However, a detailed comparison based on a selected load case is reported in D3.4. Numerous simulations were conducted by both partners which provided results in terms of kinematics, proposed HBM injury parameters and head injury risk assessment. When comparing the results, it was found that the injury risk for a subdural haematoma (SDH) predicted by THUMS-D was much lower than the SDH risk predictions by THUMS-VW. Other body regions however had comparable results, for example the maximum lateral rib deflections and pelvis forces. Larger differences were observed mainly at the higher impact velocity of 40 km/h and at the 20% offset pedestrian position for the tibia and thigh forces. Since completely identical HBMs could not be used due to the different crash software utilized differences were expected. As suggested in report D3.4 the variation of results should be reconsidered when identical THUMS models can be used in both crash codes in the future. However, the results obtained demonstrate that the entire tool chain is working properly. Each THUMS together with the SUFEHM allows a quantitative head injury risk assessment.
The analysis undertaken by CRF with the use of their Pedestrian Accident Compliant (PAC) FE dummy model in the virtual tool chain has been reported in detail as well. Similarly, results could also be compared with another model, this time in comparison to Daimler’s THUMS-D analysis on REVM1. When comparing the kinematics of the two pedestrian models, it became obvious that the PAC FE dummy shows different kinematics, mainly according to the head, which does not seem reasonable. Since the PAC dummy kinematics do not compare well with the kinematics derived from cadaver testing from the literature, the PAC FE dummy in its current version cannot be recommended to be used as part of the virtual tool chain for the assessment of pedestrian safety. A modification of the PAC FE dummy to overcome the issue of poor head kinematics was not planned and considered within this project.
When assessing pedestrian safety sensor systems in the virtual tool chain, in task T3.2 shortcomings regarding the analysed vehicle front structure of REVM1 become evident. The root cause was the weakness of the structural parts of the vehicle model. In consequence no virtual sensor evaluation was possible. Normally, state-of-the-art vehicle front ends have a stiffer structure due to e.g. strutting parts, radiator grill, etc.. This allows a better separation of “nofire” and “mustfire” signals in the virtual sensor model. An improved vehicle structure of REVM1 therefore would help as far as the application of the virtual sensor calibration approach is concerned. Hence, a virtual sensor calibration can still be a valid part of the virtual tool chain.
In addition, asphalt models have been introduced and partly used to better simulate the road rather than applying a simple rigid plane. However, no benefit could be seen in using these models when comparing head injury parameters on the HBMs. Nevertheless, the road modelling is important when the secondary impact is simulated and some kind of improved asphalt - in terms of energy absorption – will be introduced. This is planned to be further evaluated in work package WP4 of this project.
Summing up it can be concluded, that a further step is taken towards a virtual assessment of pedestrian safety using entire pedestrian models and especially HBMs. Currently, as far as specific injury risk assessments are concerned, this is limited to the head while using UNISTRA’s SUFEHM together with the post-processing IRA-tool. On the other hand, although currently being limited to the head with respect to a detailed and quantitative risk assessment, the most important body region as far as frequent severe or fatal injuries of pedestrians are concerned, can be assessed. For all other body regions further research is needed to get to model dependant injury risk assessments. The simple adaption of available injury parameter thresholds derived from tests found in the literature, e.g. a maximum plastic strain value for bone fracture, and their direct application to the post-processing of the HBMs, only allows a first rough estimation. It definitely cannot be a sufficient basis if the models are intended to be used within a virtual certification process yet. For this purpose widely accepted appropriate injury parameters and belonging model dependent thresholds do not yet exists, but are required and need to be agreed upon. Anyhow, the proposed injury parameters can be used for the optimisation in the virtual tool chain. The application of this virtual tool chain for pedestrian safety measures will be part of work package WP4 in this project.
In terms of the scattering of simulation results, the influence of small changes to the injury parameters and risk assessment could be shown by slight variation of the initial setups, i.e. position of the pedestrian relative to the vehicle. The scattering is either observed for a variation within the same model and code, but also for the same load case between the codes. Further robustness of the pedestrian HBMs as well as further improvement towards harmonised HBMs fulfilling a definite level of validation and being assessed by a kind of performance rating, is desirable. A first step to achieve a harmonised model and to define a specific validation catalogue is undertaken by the work of the THUMS User Community “TUC” where partners Daimler and VW are involved. Unfortunately, due to the timing, this first harmonised model could not yet be applied to the SafeEV project.
D3.6:
The aim of the work in WP3 of the SafeEV Project is to develop a seamless tool chain to investigate solutions for pedestrian safety and occupant protection in view of future virtual certification testing of Small Electric Vehicles (SEV). The work is based on the test conditions which were described in Deliverable 2 (D2) of the SafeEV Project. The work described in this report D3.6 concerns the work carried out in Task 3.3: Development of an advanced simulation methodology (tool chain) for a consistent safety analysis for occupant protection (incl. compatibility) in ‘Safe Electric Vehicles (SEVs). A sister report (D3.5) in SafeEV concerns a tool chain for pedestrian protection.
D3.6 is a follow-up of the report D3.3 where the objective was specifically to define requirements for such a tool chain. The main aim of the current report D3.6 is to report the work that was carried out in Task 3.3 based on these requirements in order to complete the tool-chain as well as to show some results (if applicable) to demonstrate the use of the tool-chain.
D3.6 presents tool-chains for the most important accident conditions for protection of car occupants: respectively frontal impacts, side impacts, front crash compatibility and side crash compatibility. Simulation of the occupant in the pre-crash phase as well a crash sensor modelling will become more and more important in a tool chain for the safety design of vehicles, due to the introduction of active safety systems. They will be described in separate Chapters in this report. Also post-processing is important in a tool-chain and will be described in a separate Chapter.
Both the tool-chains proposed for virtual assessment of frontal and side occupant safety respectively include a Human Body Model for the driver. The various simulations demonstrate that HBM occupant evaluation for SEV’s is possible including injury prediction.
The front and side crash compatibility research was carried out with crash dummy models: in fact, even if in the future applications HBMs are envisioned as the standard occupants within the virtual tool chain, the use of virtual dummies still will maintain in the future its relevance because of the link with the experimental crash testing. Then the inclusion of such more traditional virtual tools has been considered of importance for the development of the tool chain dedicated to SEVs.
Front crash compatibility assessment concerns questions of self and partner protection and the defined tool-chain was able to address both aspects. Also for side impact compatibility the suitability of the tool-chain was demonstrated. A compatibility index was proposed that showed a good capability to classify the level of compatibility of the SEV in both front and side compatibility. The overall behaviour of the evaluated SEV was in general adequate both for front and side impact compatibility and some possibilities to further improve the performance could be indicated.
A reliable tool-chain which involves also the pre-crash phase is important to study human response in the pre-crash phase. The objective was to study the tool-chain in an integrated pre-crash and in-crash situation. For the pre-crash phase realistic behaviour could be demonstrated in comparison with human volunteer testing but the continuation of the activated HBM in the in-crash phase caused instability problems in the HBM model (THUMS). So the current tool-chain has still limitations in the respect and therefore for the work in WP4 an alternative approach will be used.
The objective of the crash sensor modelling activity in SafeEV was the integration and assessment of frontal crash sensor system in the virtual tool-chain. The methodology could be successfully demonstrated in various crash conditions involving an SEV. For “traditional” vehicle architectures used in current small vehicles like the Smart fortwo architecture the current approach for sensor locations can be applied. In case of large changes of the vehicle architecture alternative sensor layouts have been suggested.
The pre-processing section of this report is focussing on 3 generic aspects of a crash simulation tool-chain: belt positioning, occupant positioning and simulations in a sled environment. The focus of the post-processing part is on how to deal with the vast amount of results for injury evaluation criteria produced by human body models (HBM), including the use of the Abbreviated Injury Scale (AIS).
D3.7:
The main objective of the work in WP3 of the SafeEV project is to develop a seamless tool chain to investigate solutions for pedestrian and occupant protection as a part of a virtual certification of Small Electric Vehicles (SEVs).
The work described in D3.7 concerns the work carried out in task 3.4: Revision of test configurations and evaluation criteria towards the new developed evaluation methods. This report is a follow-up of the report D2 and focuses on final test configurations (front, side and pedestrian impact scenarios) and evaluation criteria (occupant and vehicle), including the findings of WP3.
D3.7 presents a final proposal of test conditions and evaluations for virtual assessment of the most important accident conditions in urban environment related to SEVs in the future. Scenarios addressing occupant protection include front and side impacts and the compatibility of vehicles.
Having all the components of the evaluation tool chain, simulations were performed to identify the most appropriate test configuration. As the outcome, the final test configurations for the front and side impact test conditions are listed below.
The proposed frontal crash scenario, designed to asses occupant protection in SEVs, uses:
• Crash barrier: 1.300 kg pMPDB,
• Impact speed 35 km/h – 35 km/h,
• Impact angle: 30 deg
• Overlap: 50 %
• HBM in driver seat (Hybrid III or newer dummy model for comparison to experimental testing)
The following body parts are supposed to be observed:
• Head evaluation based on acceleration, stress, strain and energy
• Ribs, bones and ligaments fracture
• Heart, lungs, liver, spleen and intestines damage
The side crash scenario involves:
• Crash barrier: 1.300 kg AE-MDB
• Impact speed: 45 km/h
• Impact angle: 90 deg
• Impact point: R + 250 mm
• HBM in driver seat (WorldSID or newer dummy model for comparison to experimental testing)
The following body parts are supposed to be observed::Head evaluation based on acceleration, stress, strain and energy
• Ribs and pelvis fracture
• Ribcage compression
• Pelvic and abdomen forces
Despite the advancement of HBMs, virtual dummy models will remain in the testing protocols as a link to experimental testing.
Pedestrian protection is a vital part of the developed tool chain. As SEVs are intended to be used mostly in an urban environment, pedestrian protection evaluation is equally important as occupant protection. As a first step sensors which provide information to prevent accidents were evaluated. Secondly, in case an accident happens injury protection offered in and by the vehicle on a tissue level was analysed.
Based on this study, the final test configuration for virtual pedestrian protection is proposed as follows:
• Two velocities:
- 25 km/h to take pre-crash braking into account
- 40 km/h to simulate a non-detected pedestrian
• Impact locations with the H-point aligned to the relevant position:
- 20 % (or between wheel and fender for outstanding wheels)
- 50 % impact location of vehicle front
- 80 % (not used for outstanding wheels vehicle type)
- Directly on wheel for outstanding wheels vehicle type
- HBM pedestrian model
The following body parts are supposed to be observed:
• Head evaluation based on acceleration, stress, strain and energy
• Ribs and pelvis evaluation based on plastic strain
• Ribcage based on deflection of selected ribs
• Femur, tibia and fibula based on plastic strain
These configurations are described within this deliverable and will be used in Work Package 4 and 5 to assess the effectiveness of new generic solutions for occupant and pedestrian safety enhancement.
HBMs were used successfully in this work to numerically simulate kinematics and dynamics of occupant and pedestrian in various scenarios. Stress and strain values could be derived for various body parts. Nevertheless to fully benefit from the tool chain for a virtual certification specific injury risk curves and failure threshold values for relevant organs and individual HBMs need to be developed in future projects.
D4.1:
In D4.1 developed safety measures for reducing the injury risk for pedestrians during collisions with small electric vehicles are described. These safety measures are developed for two generic vehicle models, the so-called REVM1 and REVM2.
Following the virtual assessment approach developed in WP3, critical components for both vehicle models are defined in a first step. Afterwards solutions for each critical component are developed based on the virtual tool chain. Finally, the most promising concept, which will be further investigated in WP5, is defined and guidelines for pedestrian safety for small electric vehicles are derived.
Within the present report the potential of a stiffness adaptable CFRP is investigated. The investigated CFRP material has a thin thermoplastic coating between the carbon fibre reinforcement and the thermoset matrix resin. By applying an electric current through the carbon fibres, the thermoplastic coating is heated above the glass transition temperature, which results in reduced fibre/matrix load transfer and consequently in reduced stiffness. Due to this, the Young modulus in longitudinal direction can be reduced from 89 GPa in the stiff state to 22 GPa in the soft state.
For REVM1 safety measures for the bumper, windscreen and windscreen frame area are developed. Of special interests are the front bumper solution using a stiffness adaptable CFRP, a new approach for considering the stochastic fracture behaviour of glass and an external airbag for the windscreen frame area in an O-shape design. Furthermore, different sensor positions for improving the signal discrimination potential are considered. With the stiffness adaptable CFRP for the front bumper, the injury risks for the lower leg impactor could be drastically reduced (tibia acceleration from 302 g to 85 g, knee bending from 31 ° to 3.2 °, knee shearing from -7.1 mm to -2.0 mm). By considering the stochastic fracture behaviour of glass within the new assessment approach for laminated safety glass, the windscreen geometry could be improved and the resulting cumulated head injury criterion reduced from 805 to 577.
For REVM2 various safety measures for the lower and upper transversal beam, the lower windscreen cross beam and the outboard wheel design are developed. The usage of the stiffness adaptable CFRP is further investigated for the lower transversal and the lower windscreen cross beam. Also the effect of a foam covering the chassis of the outboard wheel design and the dependency of impact speed on injury risk are evaluated. With the stiffness adaptable CFRP used for the lower transversal beam the injury risks for the lower leg impactor could be drastically reduced (tibia acceleration from 263 g to 113 g, knee bending from 11 ° to 9 °, knee shearing from 3.5 mm to 3.8 mm). As the study will show, by integrating a foam the injury risk for lower leg fracture could be reduced until a velocity of 25 km/h. For higher impact velocities, the safety potential of foam paddings is rather low.
A further safety measure for reducing the injury risk during secondary impact on the ground is assessed based on two forgiving asphalt designs, developed by an external partner. The analysis showed that the usage of a full human body model for the assessment of injury risks during secondary impact is not feasible. Small changes during the primary impact strongly influenced the injury risk. The assessment using head models only showed a reduced injury risk for an impact on the forgiving asphalt designs.
Based on a comparison of the overall safety potential of the different measures, the lower transversal beam using stiffness adaptable CFRP showed the most promising potential, since it reduces the injury risk for a leg injury for pedestrians and might be beneficial for occupant safety during high speed collisions as well.
Finally, guidelines for pedestrian safety for small electric vehicles are derived.
D4.2:
D4.2 describes the results of SafeEV work package 4 subtask 2. The target of the subtask activities is the improvement of the occupant safety level for small electric vehicles demonstrated for the exemplary reference electric vehicle models (REVMs) available within the project.
Two types of improvements are discussed:
- Structural improvements reducing the passenger compartment intrusion level and enhancing the structural interaction capacities of the REVMs
- Optimisation of existing and the implementation of novel restraint system features with the aim to improve the interaction between the occupant and the vehicle structure in case of crash
The crash test configurations serving as design target for the vehicles are the ones defined and justified for SEV occupant protection testing within SafeEV work package 3. In addition, also further test configurations representing specific problems for the different REVMs are investigated. This results in a set of additional vehicle-to-vehicle crash configurations. In this way the discussed structural improvements can also be analysed with respect to their use potential in real world crash situations that are out of scope of laboratory testing.
The described improvements follow two targets. One is to augment the protection level for the occupants of small electric vehicles as represented by the given set of REVMs. The other is to contribute to the fulfilment of lightweight targets defined for electric vehicles where it is possible without detriment to the crash safety level.
For the improvement measures showing the best performance concerning the two targets further investigations are executed to describe additional questions such as their economic feasibility.
Having discussed all improvements for the REVMs, global design recommendations concerning the given crash test configurations summarise the activity results.
Future small electric vehicle design activities will be able to profit especially from these results in case the global structural design of their concepts is comparable to some structural aspects of the selection of REVMs. As the set of REVMs represents the state-of-the-art concerning global structural design parameters for small cars, the use benefits are expected to be high.
The structural improvement highlights at a glance:
• Aluminium space frame design front vehicle resulting in improved front crash safety and 40 % weight saving potential
• Advanced carbon fibre reinforced plastic (CFRP) material with modifiable stiffness properties applied to the front cross member resulting in balanced protection levels for occupants in case of front crash and for vulnerable road users in case of vehicle impacting
• Retractable steering wheel system allowing to enlarge the driver seat occupant survival space in front crash situations without weight-intensive structural reinforcements
• Active door beam with inflatable structure augmenting the occupant protection level in side crash situations without high design space demand
As result 14 global design recommendations are formulated concerning structural improvements to occupant safety and weight reduction for small electric vehicles.
The restraint system improvement highlights at a glance:
• Optimised driver and knee airbag behaviour increasing the safety benefit of a retractable steering wheel system for full width deformable barrier (FWDB) front crash
• Inflatable belt-bag reducing the risk of driver’s head and chest injuries in oblique mobile progressive deformable barrier (MPDB) front crash
• Pre-tensioning system for belts reducing the driver position change (out-of-position state) before frontal impact due to autonomous emergency brake system functionality
• Lateral movable seat equipped with an inflatable belt system reducing the driver’s head displacement and chest deflections and forces in side crash test configurations
Resulting from the executed investigations 8 global system design and design process recommendations are defined concerning restraint system improvements to the driver seat occupant safety.
D5.1:
D5.1 describes the results of SafeEV work package 5 subtask 1, in which a safety solution selected from WP4 is physically produced. A suitable test bench is then devised and used to evaluate the performance of the solution. As the selected component is a stiffness adaptable transversal beam for REVM2, testing is performed in both the reduced stiffness and the normal stiffness states for comparison. It is concluded that by reducing the stiffness of the material, a significant reduction in peak load transferred to the pedestrian can be achieved. Furthermore the original Crossbeam were also tested. Details of the development of the physical component and its test bench are described in this report.
Two virtual demonstrators are developed based on concepts previously discussed in WP4. These are an “O shaped” airbag that showed good potential for improving pedestrian safety, and the modelling of the restraint system configuration of REVM1.
An analysis of the REVM1 front structure updated in WP4 is performed to assess the influence of structural changes on the sensing system. A comparison is made between the design used in WP3, in which it was found the structure was too soft for state of the art sensor performance and the design from WP4. It is found that as well as increasing pedestrian safety (as shown in WP4.1) the time for activation of the pedestrian safety system is also improved.
D5.2:
D5.2 focuses on an improvement of the car front in order to increase pedestrian protection. Two versions of the car model were used to simulate a pedestrian accident in order to compare the car behaviour, the body kinematic and finally to assess comparatively the different injury risks posed. The proposed improvement is limited to the change of the front transversal beam. Therefore only lower leg injury risk was reduced. However, due to the slight changes of the body kinematic, increased head injury risk was computed. This illustrates the complexity of pedestrian protection optimization.
Even if the pedestrian protection was not significantly improved in this section, the subtask is a perfect demonstrator of the tool chain developed in framework of SafeEV.
D6.1:
D6.1 describes the intentions of the consortium members for disseminating and exploiting the project results and findings.
Within the dissemination plan strategy and planned activities are described. For these dissemination activities already possible audience and main addressees where identified within the implementation phase of the project. This report will strengthen now this information transfer in terms of a coordinated activity also within the SEAM cluster. Moreover feedback and comments on the proposed methods and concepts from relevant stakeholders and/or dedicated activities (e.g. workshops) in the course of the project will be taken into account for the formulation of the “best practice guideline” at the end of the project.
While this Dissemination Plan defines the planned activities and its coordination by the partners, the second part of this report presents a plan for using and implementing new knowledge and project findings in detail and related activities and how they support exploitation.
D6.2:
D6.2 describes the final workshop of SafeEV together with the SEAM cluster project MATISSE and attracted 46 participants. This deliverable provides details of when and how the workshop was organised and what the contents of the presentations were.
D8.1:
In the next 20 years the number of small and light-weight full electric vehicles will substantially increase especially in urban areas. These Small Electric Vehicles (SEVs) show distinctive design differences compared to the traditional car (e.g. no bonnets, vertical windscreens, outboard wheels). Thus the consequences of impacts of SEVs with vulnerable road users (VRU) and other (heavier) vehicles will be different from traditional collisions. These fundamental changes are not adequately addressed by current vehicle safety evaluation methods and regulations. Therefore it was the main objective of the project SafeEV, based on a systematic analysis and prognosis of future accident scenarios, to define advanced test scenarios and evaluation criteria for VRU, occupant safety and compatibility of SEVs. Moreover, industrial applicable methods for virtual testing of these scenarios and criteria (e.g. a method for active occupant safety assessment) were developed and demonstrated in terms of a virtual tool chain for potential extension of new normative scenario for this kind of vehicles.
Within D8.1 all project results are now reviewed and consolidated with the focus to formulate a best practice guideline for the implementation of the advanced simulation methodologies into such a “tool chain” for enhanced safety analysis concerning pedestrian and occupant safety on the one hand and also global design recommendations for dedicated and advanced safety solutions for SEVs (derived and evaluated through the application of the projected simulation methodologies) on the other hand.
Notwithstanding the main objective and above mentioned focus on virtual methods of the project the report follows now the “function chain” of an integrated safety system (traffic / test scenario > pre-crash sensing & safety system > in crash safety system). For a certain function level then the components of the virtual “tool chain” are discussed.
Best practice guide for the use of stochastic analysis and simulation to derive and predict future accident and finally test scenarios are summarized in the first part of the report. Lessons learned are discussed in terms of linking these results with other scenario and prediction methodologies.
It could be demonstrated, that virtual methods facilitated the optimization of environmental sensing in terms of scenario definition (pre-crash) and are indispensable for in-crash optimization of a crash sensor cluster (pedestrian & occupant safety) concerning interaction with vehicle and structural parts. Finally it could be stated that pre-crash sensor based autonomous braking systems should be mandatory for SEVs respectively vehicles with dedicated use in urban areas.
Structural interaction and compatibility concerning ideal load path are identified with virtual models and general design guideline are given. For this extensively the virtual tool chain including V & V (Validation & Verification) aspects are described. Especially the issue of formulating novel material models are addressed by the development of advanced and active structural components.
Safety and protection systems for pedestrians and occupants are discussed on the next functional level. In general it could be stated that any distribution of load (e.g. kneebag, inflatable belt) within the crash phase is beneficial in terms of best possible restraint effect. This also applies for advanced functions on the temporal axis. It could be demonstrated that pre-crash triggered restraint and / or pre-acceleration followed finally also by adaptive load limiting (e.g. retractable steering wheel) are effective strategies for improving the protection level offered by SEVs. Nevertheless, especially the use of virtual methods has shown, that combining such advanced systems an intelligent system management will be needed.
Beside a more standardised evaluation (a post processing tool and method was proposed in the course of the project) the projected numerical tools and models offered even more detailed information whenever specific effects were assessed or method for system parameter optimization was needed. On the other hand, it looked like, that the applied Human Body Models were quite sensitive to system parameters or vehicle design (this was observed especially for the pedestrian safety evaluation). Finally it has to be stated, that for most body regions of the applied HBM no general criteria are defined so far. However, it was possible to work effectively with the proposed injury parameters.
In the last chapter also general recommendations (e.g. comparability of simulations codes, validation of virtual models) and needs for further research are given.
Based on the results of this project, it is entirely realistic that virtual methods and tools could be implemented in evaluation and assessment procedures for specific SEVs respectively for a new class of vehicles.
Potential Impact:
Beside the common dissemination activities and in support to the communication activities of the Commission services the consortium may be requested (on an annual basis or upon request) to provide the Commission with a 2 pages information sheet which will be drafted in a standard format communicated by the Commission. The Commission services may also request on illustration (picture, schema or drawing) to illustrate such communication material.
A first project information was requested by the Commission in October 2013 for the edition of the SST projects synopsis, which was responded accordingly. The project is also asked to contribute to relevant dissemination, international cooperation and information exchange activities of interest to the Commission such as those within the scope of the European Green Cars Initiative. Also for this EGVI dissemination area already a contribution was prepared. SafeEV has provided a slideshow which overviews concisely their objectives and major outcomes to the European Green Vehicles Initiative Association for the TRA 2014 in Paris. The key audience and main addressees of dissemination activities are already defined in the Description of Work respectively during the implementation phase of this project. Nevertheless, this list will be updated continuously in the course of the project and will follow also feedback from dedicated and cooperative activities (SEAM workshops, network activities etc.).
Most of the activities started in month one. Five main categories were identified to build the core of the dissemination strategy of this project.
Project oriented dissemination & information platform
Just after launch the project SafeEV had its own website (www.project-safeev.eu) making project information available online. It was further developed during the first project period and will be regularly updated throughout the project. The site is interactive and contains project and partner description. It also builds the “first point of contact” for any interested parties which intend to get in contact with the consortium or searching for results, information to download or further links.
For this a newsletter was created, which will be send regularly to registered recipients. The registration can be done via this website.
Also all public reports listed on the website and will be made available by direct download. Finally also direct links to the SEAM cluster and the individual SEAM projects are established on the website.
Scientific publications and contribution to conferences
The participation in international conferences and contribution to scientific publications makes it possible to communicate SafeEV objectives and findings to wide expert public. The academic partners will especially contribute and prepare publications to conferences and journals. Target conferences are e.g. SAE, ESV. Most of the (scientific) partners in the consortium are closely and well involved in the scientific community or even members of a programme committees or editors of mentioned journals which will make it possible to submit a number of specific publications and issues on project findings and results.
This list follows the dissemination strategy and further road map for publications below:
First year dissemination
- Generate attention for SafeEV to wide expert public
- Communication of project objectives and tools & methods which will be used
Second year dissemination
- Communication of first project results (scenarios / prognosis → basis for further activities).
- Demand information & feedback from expert public and related initiatives / research
Third year dissemination
- Ongoing status report / results from project core activities
- Communication of first project findings & recommendations
Final / Fourth year dissemination
- Final project results and recommendations
- Provide guidelines to expert public and other stakeholders
- Initiate follow up activities (including SEAM Cluster)
- Initiate exploitation of results (including SEAM Cluster)
Implementation of project findings – interest groups & relevant stakeholders
Mainly the industrial partners (Daimler, CRF, VW, Pininfarina, Bosch) will initiate and coordinate cooperative activities in the project phase when it comes to implementation, demonstration and exploitation of the project results. This will be done via communication and direct transfer of project findings to interest and working groups (e.g. TUC, ACEA, VDA, CLEPA). Also relevant stakeholders in the area of vehicle and traffic safety (e.g. NCAP, EEVC, UNECE) will be addressed via dedicated activities.
Similar to the communication via the scientific community (3.2) the current network activities of the (industry) project partners will be used as a first level dissemination channel.
SafeEV activities are tracked in the TUC – THUMS USER COMMUNITY (www.tucproject.org). Focus is especially on the discussion and definition of dedicated methods for post processing of HBM simulation results and injury risk assessment. TUC represents a group of THUMS user, namely Audi, Autoliv, BMW, Daimler, Opel, Porsche, Toyota, VW and LMU (Ludwig Maximilians University Munich) as coordinator, which initiated this platform (Collaborative Project) end of 2012 to set up a framework and harmonize general and administrative requirements for the implementation of FE- Human Body Models in vehicle and traffic safety applications.
SafeEV activities are also communicated to and within PDB (www.pdb-org.com) - Partnership for Dummy Technology and Biomechanics, which was founded in 2002 as a cooperative venture between German car manufacturers in the competitively neutral field of crash test dummy technology, biomechanics and simulation. Since Feb. 2014 a sub-group addressing the use and general implementation of virtual HBM to development and assessment methods is installed in this organisation.
A direct link via a project partner also exists to the EEVC –European Enhanced Vehicle-Safety Committee and its WG 12 “Crash Dummies”. Another relevant EEVC WG in terms of exploitation of project findings is WG 22 “Virtual Testing”.
Beside these direct transfer of project results by personal / partners contact workshop events are seen as most appropriate activity.
The exploitation strategy of this initiative is mainly determined by the specific character and content of the research and project objectives.
On the one hand the main objective of this project is to develop and demonstrate a complete new assessment methodology for a vehicle class (or mobility mode), which up until now are not subject of regulatory or consumer testing requirements. So, ideally, stakeholders in this area might take offer the complete approach respectively methodology, tools and criteria (“tool chain”).
For this, this core result could be seen and defined as a pilot for a new generation and widely accepted assessment method. The basis for such a pilot is already provided by a long term process for which IMVITER delivered latest and relevant requirements. This general process for implementation of Virtual Testing in safety regulations and/or assessment was initiated 2006 by the CARS21 initiative and continuously further developed within research projects and even the EC regulatory framework.
Figure 4.1 illustrates this background and the optional continuation by a “SafeEV pilot”.
Figure: Steps in the implementation of VT in safety regulations and – assessment and context of a “SafeEV pilot”.
On the other hand, SaveEV is also providing all relevant components respectively “chain links” for a virtual testing environment & tool chain needed for the evaluation of pedestrian & occupant protection and increased compatibility of SEVs. So, also exploitation of partial results and specific findings will be promoted by project activities and even more by exploitation plans of the partners. The next paragraph highlights especially these more specific areas and dedicated project findings.
List of Websites:
http://www.project-safeev.eu/
The project SafeEV, “Safe Small Electric Vehicles through Advanced Simulation Methodologies”, was funded by the European Community’s Seventh Framework Programme under the Transport programme (Grant Agreement no: 314265). It consists of 11 partners from 5 different countries. Our consortium combines industry (four OEMs, one TIER 1 supplier, one SME) and research (five research organisations) in the field of vehicle safety.
The number of small and light-weight full electric vehicles will substantially increase especially in urban areas. These Small Electric Vehicles (SEVs, in SafeEV in particular concern vehicles in the mass category of so-called L7e vehicles) show distinctive design differences compared to the traditional car (e.g. no bonnets, vertical windscreens, outstanding wheels). Thus the consequences of impacts of SEVs with vulnerable road users (VRU) and other (heavier) vehicles will be different from traditional collisions. These fundamental changes are not adequately addressed by current vehicle safety evaluation methods and regulations. VRU protection, compatibility with heavier opponents and the introduction of active safety systems have to be appropriately taken into account in order to avoid any SEV over-engineering (e.g. heavy or complex vehicle body) by applying current regulations and substantially impair the SEVs (environmental) efficiency.
Based on future accident scenarios, the project SafeEV aims to define and develop advanced test configurations and evaluation criteria for VRU protection, occupant safety and compatibility of SEVs. Moreover, industrial applicable methods for virtual testing of these configurations and criteria (e.g. a method for active occupant safety assessment) will be developed. These methods are applied in order to derive protection systems for enhanced VRU and occupant safety for SEVs. The evaluation of one developed hardware system was used to demonstrate the potential and applicability of these new methods. Dedicated best practice guidelines for VRU and occupant safety evaluation of SEVs will ensure a sustainable impact for industry and regulatory organisations beyond the project duration. With the new evaluation methods developed, vehicle safety for SEV on urban roads in the next decade will be adequately addressed resulting in less fatalities and injuries without compromising vehicle efficiency. Moreover cost-efficient development of SEVs will be made possible by the new virtual testing methodologies developed.
Project Context and Objectives:
WP 1:
The objective of the work package is the identification of the most relevant future accident scenarios involving small electric vehicles (SEVs). These SEVs are coming to the road in the near future and therefore prediction of the most relevant accident scenarios of these vehicles has to be made. SafeEV has focused on the crashworthiness of SEVs looking into mitigation of the crash-severity during pre-crash and crash. A focus is set on the influence of pre-crash safety systems on future (next decade) accident scenario’s as well as on pedestrian (vulnerable road user) protection and compatibility for future small electric vehicles. (pre-crash and compatibility are addressed in Matisse project and transferred to SafeEV).
Two Tasks were defined within WP1:
• Task 1.1 Methodical analysis on future pedestrian accident scenarios involving SEVs (M1-6)
• Task 1.2 Conclusions on future accident scenarios including results from MATISSE (M7-8) [Input from MATISSE]
Task 1.1 Methodical analysis on future pedestrian accident scenarios involving SEVs (M1-6)
Summarized the main assessment steps for the prediction on future accident scenarios is summarized with the following bullet points.
• Accident Type Assessment
• Efficiency Analysis Method for Driver Assistance Systems
• Stochastic Accident Prediction Approach
• Delphi Study (Expert Survey)
• Public Survey
• Pedestrian Kinematics Analysis including New Car Geometries
Task 1.2 Conclusions on future accident scenarios including results from MATISSE (M7-8) [Input from MATISSE]
Task 1.2 was to summarize the most important findings of the Matisse D1.1 and the SafeEV D1.1 documents.
Matisse D1.1 was focused on occupant related scenario prediction. The study also included a review of current Swedish car accidents and their relevance for future scenarios in view of forecasted safety systems developments. In addition accident data from the German In-Depth Accident Study (GIDAS) database were extracted. Smaller M1 vehicle models were categorised as Small Urban Vehicles and compared to a matched set of heavier M1 vehicles. Only accidents that occurred in conditions that would be likely for a small urban vehicle were included (maximum 2 occupants, maximum speed 100 km/h, maximum travel distance 100 km).
WP 2:
WP 2 “Advanced test configurations and criteria for pedestrian and occupant safety in SEVs” comprised three tasks:
Task 2.1: Definition of test conditions for vulnerable road user protection in crashes involving small electric vehicles
Task 2.2: Definition of test conditions for occupant protection (incl. compatibility) assessment in crashes involving small electric vehicles
Task 2.3: Criteria for assessment of SEV safety in crashes (M9-12)
The objectives of WP2 are:
• Development of necessary test configurations (accident conditions) for vulnerable road user and occupant protection assessment (incl. compatibility) in accidents involving SEV’s.
• Specification of the methodologies and tools for safety assessment in the proposed test set-ups (experimental, virtual as well as hybrid approaches).
• Assessment of the applicability of current evaluation criteria for SEVs and promote the development of new criteria where current criteria are inadequate to address SEVs safety. This concerns both vehicle based criteria incl. fire and electrical safety criteria and injury assessments criteria. The criteria should be suitable for virtual testing.
The test conditions are based on the results of WP1 of SafeEV, recent regulatory trends and new developments in the field of consumer testing (NCAP). The assessment of pre-crash based injury reduction systems is also taken into account. The work included specification of the (virtual) tools to be used like impactors, virtual models, crash dummies, integrated experimental/virtual methods, as well as criteria to be used for the assessment. A first proposal of test scenarios for frontal and side impacts (occupant protection) as well as a proposal for pedestrian and cyclist protection for SEVs in urban environment were developed.
WP 3:
Work package 3, Advanced simulation methodology for integrated pedestrian / occupant safety in SEVs, overall goal is to develop a seamless tool chain to investigate solutions for pedestrian safety and occupant protection towards virtual certification testing. Therefore the following objectives will be pursued:
• Develop REVMs for the development of the advanced evaluation methodologies
• Develop simulation and testing tools for pre-crash and crash safety evaluation needed for the test configurations and criteria from WP2
• Develop tool chain for pedestrian safety analyses (based on existing human models which will be enhanced within this project by use of brain model, active safety evaluation, ...)
• Develop tool chain for occupant protection incl. compatibility analyses (using e.g. generic vehicle fleet on the test configurations and criteria from WP 2)
• Develop pre-crash sensing models and integrate them into virtual vehicle models of SEVs in order to create virtual pre-crash-data and evaluate feasibility of optimized sensing technologies and topologies.
• Final revision of the test configurations and criteria from WP2
The work package is splitted into four tasks
Task 3.1: Setup of two reference electric vehicle models (REVM)
Task 3.2: Development of an advanced simulation methodology for consistent safety analysis for pedestrian protection for SEVs
Task 3.3: Development of an advanced simulation methodology for a consistent safety analysis for occupant protection (incl. compatibility) in SEVs
Task 3.4: Revision of test configurations and evaluation criteria towards the new developed evaluation methods
This work package was started in project month 4 (January 2012). The first task was the setup of two reference electric vehicle models (REVM). These two models are essential for the whole simulation work, because these are used for further activities for the development of the advanced evaluation methods. Two different models were generated: one, the REVM1, was conceived since the beginning as a native electric car while the other, the REVM2, was derived from an existing ICE small vehicle. Both REVMs were originally built within LS Dyna F.E. code; subsequently, only the REVM2 was translated into PAM-Crash code, too. Task 3.2 and 3.3 were focusing in the development of an advanced simulation methodology for pedestrian and occupant safety and define a virtual tool chain, results are summarized in deliverable D3.5 and D3.6. Within D3.7 the revision of test configurations and evaluation criteria towards the new developed evaluation methods, based on D2 and updated with findings of WP3.
WP 4:
Work package 4 is dedicated to the definition of generic safety solutions to improve the safety level of SEVs with respect to occupant and pedestrian protection and the assessment of the defined measures applying SafeEV’s virtual evaluation tool chain. Two subtasks are defined for work package 4:
Task 4.1: Definition and assessment of advanced pedestrian safety measures for SEVs.
Task 4.2: Definition and assessment of advanced occupant safety measures for SEVs.
The objectives of the two tasks within work package 4 are as defined:
• Definition of generic improvement measures to augment the safety level of SEVs with respect to occupant and pedestrian protection
• Application of the virtual evaluation tool chain defined within SafeEV project work package 3 to investigate the effect of the respective measures on the fulfilment of the crashworthiness target criteria defined for SEVs as represented by the REVMs used within the project
• Identification and more detailed elaboration of the most promising solutions focusing on the crashworthiness level and lightweight targets for SEVs
• Formulation of guidelines for occupant and pedestrian protection for future SEV design activities summarising the lessons learnt during the execution of the different study steps within work package 4
All planned objectives have been fulfilled.
The results of task 4.1 are documented within SafeEV project deliverable no. D4.1 “Generic safety solutions and design candidates evaluated with the advanced methodology for pedestrian safety”. The reporting of the results of task 4.2 can be found in SafeEV project deliverable no. D4.2 “Generic safety solutions and design candidates for occupant protection including compatibility improvements”. In general it can be stated that various generic safety solution were performed and the most promising component were identified. This component is then be evaluated within WP 5.
WP 5:
The workpackage is titled, “Use case for enhanced safety for SEVs using virtual testing” and is split into two subtasks:
Task 5.1 Selection of the use case and dedicated design (M31-34)
Task 5.2 Evaluation of the use case (M34-35)
This work package elaborates a use case for an advanced safety solution and applies the evaluation methodology developed in this project. In T5.1 a pedestrian protection solution investigated in WP4, a stiffness modifiable transversal beam, has been selected for further development and physical prototyping. The manufactured beams have been produced and small scale versions tested as well as the originally beam. Furthermore, two more REVM1 concept solutions are built up to the level of virtual demonstrators and the front structure of REVM1 has been analysed with respect to signals. In T5.2 the virtual version of this beam has been included in the FE model and used to perform the virtual testing.
WP 6:
The main objective of this work package was to promote dissemination and exploitation of the project findings and results through organisation of dedicated workshops and information exchange activities and presentation at relevant conferences and scientific publications. One main task is formulated in Annex 1 of the project for WP6 which addresses dissemination and exploitation activities and the role of the partners in this area. Most of the activities started in month one. Five categories have been processed:
• Establishing project oriented dissemination & information platform (website, leaflet, etc). (in agreement and coordination with WP8 / SEAM activities)
• Contribution to and publication via conferences and scientific journals
• Implementation and exploitation of the main project findings to interest groups and relevant stakeholders outside the consortium.
• Deliverables (public) – as reference for further & subsequent activities
• Dedicated workshops (Final dissemination workshop)
WP 8:
The main objectives are communication within the project (meetings, reviews, decision making process, technical project management), the coordinate activities within the SEAM cluster and to provide the best practice guideline as one of the major project outcomes.
The work package is substitute into 3 tasks. In task 8.1 all project results will be reviewed and consolidated aiming at the formulation of a best practice guideline for the implementation of the advanced simulation methodologies into and “tool chain” for enhanced safety analysis concerning pedestrian and occupant safety in SEVs.
Task 8.2 comprises all activities within the SEAM cluster (SEAM project office, liaison team and dissemination strategy, starts in month one and ends in month 36. The SEAM cluster was initiated in order to coordinate and harmonise the four projects SafeEV, ENLIGHT, ALIVE and MATISSE. Main purpose of the SEAM cluster is to realise and monitor synergies between the four projects on RTD and demonstration level and to execute joint dissemination and exploitation activities.
Task 8.3 is the technical project management and communication and infrastructure for quality monitoring, decision making and conflict resolution.
Project Results:
D1.1:
Within Task 1.1 of WP1 most relevant future pedestrian accident scenarios involving small electric vehicles had to be estimated. This directly addresses next decade small electric vehicles (SEV´s) which will be seen in urban areas. The methodology used to identify the scenarios includes four main steps. Analysing national and in-depth accident databases give an overall scenario definition and the boundaries for a generic accident type. Stochastic simulations of the defined scenario show a trend of future impact speed. A Delphi study, supported by a public survey, estimates future mobility and city-layouts and helps to identify boundary conditions. And finally the fourth topic pedestrian kinematic analysis to evaluate the effect of future vehicle shapes and precrash systems.
From the database analysis six accident scenarios where seen as most relevant for future urban traffic, which includes near- and farside accidents before/after and outside of junctions (vehicle going straight). A generic simulation model is built up for a stochastic analysis, where the results show an overall decrease of speed with a pedestrian protection system on-board. The results of the Delphi study confirmed the assumptions made for future traffic scenarios. The consulted experts predict a decrease of average vehicle weight with a higher share of SEV´s and a dynamic reduction of future speed limits. Furthermore SEV´s being produced after 2017 will be equipped with enhanced safety features which will provide better pedestrian protection. Public transport will be improved and will have higher shares. The demographic change and the trend of aging in cities will also influence accident scenarios, but should be partly solved with an improved traffic management. The public survey shows similar trends with only a few deviations (e.g. estimated higher share of driver assistance systems). In the pedestrian kinematics analysis four future SEV designs are compared with conventional passenger cars with the focus on head contact including impact and post impact phase. It is shown that the impact speed of an adult and child head reduces with a high share of windscreen contacts due to vehicle shapes. Thus, a reduction of impact speed for future pedestrian accidents is expected. Most frequent predicted velocities will be between 10 km/h and 30 km/h.
D1.2:
D1.2 summarizes the most important findings of the Matisse D1.1 and the SafeEV D1.1 documents. The two projects are part of the SEAM cluster and they collaborate on the analysis and prognosis of future accident scenarios involving small alternatively powered vehicles (APV) in urban areas.
The methodology of the Matisse D1.1 and the SafeEV D1.1 included the following steps: 1) Analysing national and in-depth accident databases give an overall scenario definition and the boundaries for a generic accident type. 2) Stochastic simulations of the defined scenario show a trend of future impact speed with a generic pre-crash system on-board. 3) A Delphi study, supported by a public survey, estimates future mobility and city-layouts and helps to identify boundary conditions. 4) Review of current heavy truck accidents and their relevance for future scenarios in view of forecasted developments of safety systems. 5) Identification of critical accidents for high-pressure storage tanks by expert brainstorming and finite element methods. 6) Pedestrian kinematic analysis to evaluate the effect of future vehicle shapes and pre-crash systems. Four future SEV designs were compared with conventional passenger cars with the focus on head contact including impact and post impact phase.
The present report also included a review of current Swedish car accidents and their relevance for future scenarios in view of forecasted safety systems developments. In addition accident data from the German In-Depth Accident Study (GIDAS) database were extracted. Smaller M1 vehicle models were categorised as Small Urban Vehicles and compared to a matched set of heavier M1 vehicles. Only accidents that occurred in conditions that would be likely for a small urban vehicle were included (maximum 2 occupants, maximum speed 100 km/h, maximum travel distance 100 km).
Conclusions, Car to Car Accident Scenarios:
The stochastic analysis results show an overall decrease of the number of crashes and a decrease of the collision speed. For future cars of lower mass, crash delta-v will however increase. The results of the Delphi study predict a decrease of average vehicle mass with a higher share of APV´s and a dynamic reduction of future speed limits. The public survey shows similar trends with only a few deviations (e.g. estimated higher share of driver assistance systems).
Conclusions, Car to Pedestrian Accident Scenarios:
Six pedestrian-to-car accident scenarios where seen as most relevant for future urban traffic, which includes near- and far side accidents before/after and outside of junctions (vehicle going straight). A generic simulation model using stochastic analysis show an overall decrease of collision speed with a pedestrian protection system on-board. The results of the Delphi study confirmed the assumptions made for future traffic scenarios. A higher share of SEV´s and a dynamic reduction of future speed limits was forecasted. Furthermore APV´s being produced after 2017 will be equipped with enhanced safety features which will provide better pedestrian protection. It is shown that the impact speed of an adult and child head reduces with a high share of windscreen contacts due to vehicle shapes. Thus, a reduction of impact speed for future pedestrian accidents is expected. Most frequent collision speeds are forecasted to be in the range 10 km/h to 30 km/h.
In the present work several different and independent data sets and methodologies have been applied. They all include significant uncertainties. It is thus reassuring to be able to conclude that they all forecast similar trends. This gives the results a significant robustness.
In summary, it becomes obvious that certain collision types will be reduced in numbers and in average collision severity, but in all cases an estimated residual of likely 10% or more will remain and there will be a significant remainder also of severe collisions. The general publichave become used to the current vehicle safety level and will have very limited acceptance for increased injury risks as a result of the downsizing of the car fleet. If people lose trust in the new generations of small light vehicles they will refuse to buy them and they will penalise politicians that try to enforce their introduction. It must therefore be concluded that SEVs or APVs developed for production in 2025 must meet at least the same occupant and pedestrian crash protection requirements as for current vehicles
D2:
The objective of D2 is the specification of test configurations for vulnerable road user and occupant protection assessment (incl. compatibility) in accidents involving small EV’s in urban areas. Small EV’s in SafeEV in particular concern vehicles in the mass category of so-called L7e vehicles, where currently hardly any safety requirements exist.
The test conditions are based on the results of WP1 of SafeEV, recent regulatory trends and new developments in the field of consumer testing (NCAP). The assessment of pre-crash based injury reduction systems is also taken into account. The work includes specification of the (virtual) tools to be used like impactors, virtual models, crash dummies, integrated experimental/virtual methods etc...
Concerning evaluation of pre-crash and active safety systems, discussed in Chapter 2, it was concluded that the future focus in SafeEV will be on passive safety pedestrian and crash sensors and the according requirements for sensor evaluation tests. For pedestrian safety sensors it was shown that through their geometry SEVs pose a challenge on the setup of the sensors as well as on the sensor and trigger time of in-crash safety systems for pedestrians. Further it was concluded that only simulation/virtual tools will allow a continuous observation and evaluation of integrated safety systems in terms of a benefit based assessment.
For VRU (i.e. pedestrian and cyclist) safety evaluation in SafeEV, presented in Chapter 3, simulations are proposed using human body models (HBM) in the following 4 sizes: 6 year old child, 5% Female, 50% Male and 95% Male. Simulations concern impacts of a pedestrian against the vehicle front at 2 speed ranges: 25 km/h to 30 km/h (lower boundary) and 45 km/h to 50 km/h (upper boundary). Also simulations for cyclist protection are proposed.
Occupant protection is discussed in Chapter 4. The following accident types have been addressed: frontal, side, rollover, rear, compatibility and Multiple Impact Crashes (MIC’s).
For frontal occupant protection for L7e class vehicles a test (simulation with HBM) with an oblique impact configuration (30 degrees) is proposed. One option here is that also accounts for compatibility aspects, is a test with at a test speed of 30-40 km/h against a movable barrier that represents the average opponent vehicle mass (1125 kg in GIDAS analysis). Such a configuration also adapts automatically the effect of variations in SEV mass. The option for a full virtual test method here has high potential due to the fact that for crash barriers already validated models are available. Verification and Validation issues concerning the vehicle model are in general addressed by the IMVITER project and should be taken into account.
Concerning future frontal compatibility protection of M1 vehicles the ODB offset test configuration proposed by the FIMCAR project as well the FWDB test, are identified as most suitable tests to be used within the SafeEV project. The quantification of the use potential of a test configuration using a movable barrier for compatibility tests addressing SEV-to-SEV and SEV-to-M1 vehicle front crash behaviour will be an aspect addressed within the project work package 3.
For side impact occupant protection in L7e vehicles a test (simulation with HBM) using the latest MDB with a mass of 1100kg reflecting typical future car masses, is proposed. The barrier will hit the vehicle with a speed of 40 km/h at an angle of 90º. The impact location is moved backwards reflecting a short bonnet in a car body.
Rollover protection is not an important accident scenario in Europe (compared to the USA). Within SafeEV it is proposed that the SSF (Static Stability factor) and AST (Airbag standing Time) are calculated. For this no simulations or tests are needed. Eventually if these values exceed certain critical values it could be considered to perform so-called Fishhook and ESC tests which can be done by simulations. An Inverted Roof Crush test could be considered as well but will not be addressed within SafeEV.
Concerning rear impacts the introduction of AEB and introduction of seat concepts aimed at lowering the risk of WAD (Whiplash Associated Disorders) will reduce the number and risk of WAD in future rear-end impacts. However WAD will remain frequent and the consequences will continue to lead to large societal costs and personal suffering. Additional and improved test tools and sled test conditions are under development. Among these are a female dummy and associated limits and an adult 95Th percentile dummy. Also there is potential for simulations with HBM’s. Initial positioning of the occupant is a parameter having a large influence.
A final accident condition that was considered are the Multiple Impact Crashes (MIC’s), in which a vehicle experiences at least 2 impacts after each other. Although the frequency and injury severity of MIC’s will decrease due to the trends presented in WP1, a strong need to protect people in MIC’s remains due to the large frequency of these crashes and the relative high injury risk.
Recommendations concerning future work in SaveEV (WP3 etc...) concerning MIC’s are: (1) Inclusion of occupant simulations (virtual testing) with active HBM of extreme controlled vehicles manoeuvres that are aimed to reduce the risk of or the severity of second (or more) impacts in a MIC. The aim of such simulations is to study the effect on human position changes during such manoeuvres and the resulting risk on OOP’s and (2) Inclusion of occupant simulations aiming of optimization of reversible and irreversible restraint systems during MIC’s.
Chapter 5 deals with criteria to be used in the various test (simulation) conditions. A distinction is made between injury criteria, compatibility criteria and fire and electric safety criteria.
The overview of injury criteria is rather extensive. Most important body parts and future trends are addressed (including the need to have better criteria for children and elderly). Both global criteria (criteria that also can be determined at physical crash dummies) and criteria on tissue level to be determined on human body models are described. Also a distinction is made between criteria for vulnerable road users (pedestrians and cyclists) and car occupants.
The compatibility evaluation criteria defined within the FIMCAR project for the FWDB and the ODB test are initially identified as also suitable for the assessment of small electric vehicles. But it should be realized that FIMCAR didn’t integrate a force compatibility assessment criterion into the final metric proposal, so this may need some additional attention.
Concerning fire and electric safety criteria an overview of current regulations for crash safety of Battery Electric Vehicles (BEV) is given. All requirements are for an experimental evaluation procedure, focusing on frontal, side and rear impacts. A virtual assessment of these criteria is not considered yet. Possible reasons might be the limited knowledge of or the experience with the crash safety of BEV and the lack of suitable simulation models. A review of trends in this field did not predict big changes in the current requirements for 2025.
D3.1:
D3.1 describes the Reference Electric Vehicle Models (REVMs) with respect to their requirements, capabilities and validation status.
These REVMs are used within the Project as generic analysis tools for the development of those advanced evaluation methodologies that will form the pillars of a virtual tool chain to be used in future virtual certification procedures/protocols specifically dedicated to small electric vehicles, an aspect that represent one of the main goals of SafeEV.
Such models are intended to represent possible small electric vehicles including solutions that will be distinctive for the next generation of SEVs, i.e. new designs which are quite different from current traditional cars. New concepts of Small Electric Vehicles for urban areas (short SEVs) already presented to the public might belong to the M1 category (motor vehicle for passenger transport with not more than eight seats excluding driver’s seat) or to the L7e category (four-wheeled vehicle up to 400 or 500kg). Then, the approach followed in REVMs generation was to model very small M1 category vehicles that can be used even as representative of L7e category, through an adequate and artificial reduction of the total vehicle mass. In fact, this is a possible and practical option especially for those numerical simulations where pedestrian impact configurations are involved.
Two different models were generated: one, the REVM1, was conceived since the beginning as a native electric car while the other, the REVM2, was derived from an existing ICE small vehicle. Both REVMs were originally built within LS Dyna F.E. code; subsequently, only the REVM2 was translated into PAM-Crash code, too, due to specific operational needs of T3.2 related to pedestrian impact simulations with human body models. The details about such model translation are described within deliverable D3.4.
These models can be used to evaluate the behaviour of the crash structure (e.g. crash pulse, deformation characteristics and intrusions) in typical car-to-barrier and car-to-car impacts.
Moreover, they can be used for the simulation of pedestrian impacts (through pedestrian impactors and standing pedestrian dummies and/or Human Body Models).
However, no restraint systems are included in these models thus no assessment of occupant readings is possible with these first releases of REVMs: such (generic) restraint systems will be added in a subsequent phase of the project (T3.3) in order to perform the integrated biomechanical/structural work required when studying compatibility aspects of SEVs.
In any case, for the preliminary assessment of the occupant loading conditions with the first REVM releases described in this deliverable, the evaluation of the crash pulse obtained from their numerical crash simulations remains possible, by the application of known severity indicator like the OLC (Occupant Load Criterion).
The simulation runs performed so far showed that the REVM models are numerically stable and behave in a realistic way.
D3.2:
WP3 of SafeEV deals with an advanced simulation methodology for integrated pedestrian / occupant safety in small electric vehicles (SEVs). D3.2 is part of Task 3.2 of work package 3 and concerns the development of an advanced simulation methodology for consistent safety analysis for pedestrian protection for SEVs. Task 3.2 runs in parallel with Task 3.3 which deals with the corresponding development of a simulation methodology, but for occupant protection only. The objective of the D3.2 report is as an initial definition of requirements for consistent safety analysis for pedestrian protection in SEVs. These requirements are interpreted as the necessary steps to develop a so called “seamless tool chain” in order to virtually assess and optimise pedestrian safety. The main aspects covered by this report are the following:
• test conditions and body regions to be evaluated (Chapter 2)
• simulation tools to be used by the different partners (Chapter 3)
• a brief look back to the EC funded project IMVITER (IMplementation of VIrtual TEsting in safety Regulations) in order to provide an exemplary process and protocols which could enable type approval through virtual testing (Chapter 4)
• description of the necessary steps and tools needed to develop a “seamless tool chain” for pedestrian safety (Chapter 5) consisting of:
- agreement on how to evaluate injury risk using human body models (HBMs)
- comparison of a HBM vs. a pedestrian accident compliant (PAC)
- code dependence in the application of Finite Element (FE) HBMs
- utilisation of an advanced ground model to evaluate the secondary impact
- comparison of acceleration sensor signals derived from pedestrian impactor and HBM contact against the vehicle
The reporting of Task 3.2 will be completed by three future reports. Report D3.4 will make a comparison of the simulations with the models running under different FE codes for identical load cases. Thereafter, report D3.5 will describe the resulting methodology for the virtual tool chain for pedestrian safety simulation. This will include an analysis of the robustness of the virtually gained results by slightly varying the defined load cases. Finally report D3.7 will summarise the final definition of the relevant load cases and appropriate criteria for injury risk evaluation using HBMs as a result of a separate Task 3.4 in WP3. Hence report D3.7 finally closes Tasks 3.2 to 3.4 and WP3 itself.
D3.3:
The work in WP3 concerns the development and evaluation of virtual test methods and simulation tools for small electric vehicles. D3.3 is the first deliverable of task 3.3. The general objective of this report is to define the requirements for the development of an advanced simulation methodology for a consistent safety analysis for occupant protection (incl. compatibility) in SEVs.
The aim of this report is to specify:
• the test conditions to be evaluated in task 3.3 (Chapter 2);
• the simulation tools to be used in task 3.3 (Chapter 3);
• the characteristics that a protocol for type approval through virtual testing shall have, according to the work developed by the IMVITER (IMplementation of VIrtual TEsting in safety Regulations) project (Chapter 4); and
• the description of tools that will be developed in the remaining part of task 3.3 to complete the virtual tool chain (Chapter 5).
The report includes the simulation test matrices which states which accident scenarios will be evaluated and by whom. Other important information is the specification of which criteria to use in the evaluation. The matrices and criteria are specified in Chapter 2. The criteria are not finally concluded, but an important activity of task 3.3 is to further develop and agree on relevant criteria. Based on these criteria, needs to develop suitable post-processing tools will be confirmed and/or arise. All these mentioned activities will be worked on jointly and in parallel by the partners. A description of who does what in general terms is provided in Chapter 5. The final results of task 3.3 will be included in the D3.6 report of SafeEV.
D3.4:
WP3 of SafeEV addresses the development and demonstration of a seamless tool chain in order to virtually develop, optimize and assess integrated occupant and pedestrian safety in small electric vehicles. Within Task 3.2 the main focus is on the development of an advanced simulation methodology for consistent safety analysis for pedestrian protection and the implementation of WP2 scenarios and evaluation criteria. FE- Human Body Models (HBM) are identified as possible candidates for such an application and already described in detail in deliverable D3.2 (pedestrian). Finally the HBM THUMS and also the REVM2 (see D3.1) are used by two project partners in similar test scenarios applying similar criteria – but utilizing different FE solver, namely LS-Dyna and PAM-CRASH/VPS.
In terms of a reliable and universal tool chain it is essential that these simulations and models generated under different codes deliver comparable results concerning a dedicated criterion – in this case injury risk prediction.
D3.4 will therefore explicitly discuss the code dependency, translation and finally harmonisation aspects in the application of FE-HBM and the projected REVM.
In the first part of the report simulation results, generated under identical load cases, of THUMS-D (Ls-Dyna) and THUMS-VW (VPS) are compared. Following recommendations from AMSE concerning V&V (Verification & Validation) this was exemplarily done on different discretisation levels of the HBM: Component level (here: Tibia & Femur) > body region level > full body level / kinematics. Also the implementation of SUFEHM (Strasbourg Univ. Finite Element Head Model - incl. post-processing tool) under different codes and the development of code specific criteria are discussed within this chapter.
The second part of the report addresses the code dependency and translation aspect of the vehicle model REVM2. Results from identical load cases (impactor tests) were compared for REVM2 running under LS-Dyna and the REVM2 under VPS.
The comparison shows, that especially the vehicle model in the two codes delivers quite comparable results when standard impactors are used.
Even though the THUMS models were not “similar” in this study, it could be stated, that the general injury risk prediction given by the two models is quite comparable. SUFEHM, as “identical” body part, has shown similar risk prediction for two injury mechanisms. Nevertheless, and this might be the reason for the discrepancy concerning the third injury mechanism, the head contact is also influenced by slightly different full body kinematics.
Further on a comparison of simulation results in terms of comparable prediction of injury risk will also be done with the PAC model and the REVM 1 w.r.t THUMS-D in the same impact condition. This will be reported in D3.5.
D3.5:
Work package WP3 deals with the “Advanced simulation methodology for integrated pedestrian / occupant safety in Small Electric Vehicles (SEVs)”. The work package is split into four tasks and mainly describes the development of the virtual tool chain for occupant and pedestrian safety evaluation.
D3.5 is a follow-up on the D3.2 report which defined the requirements of the models in the virtual tool chain to be applied to the test setups for the pedestrian protection evaluation as defined in report D2. Hence, the report D3.5 completes the documentation of the work carried out under task T3.2 “Development of an advanced simulation methodology for consistent safety analysis for pedestrian protection for SEVs”.
The main objective of this deliverable D3.5 is to report on the application of the selected models within the virtual tool chain for pedestrian protection. First simulations of the relevant load cases have shown that further specifications of boundary conditions for the simulations became necessary. These are documented in the introductory chapter of this report along with descriptions of the updates as for the vehicle models, the Human Body Models (HBMs) and the Strasbourg University Finite Element Head Model (SUFEHM). The head injury risk assessment based on UNISTRA’s SUFEHM was enhanced during the course of task T3.2 by the extension of implied real world head trauma data together with an update of the Injury Risk Assessment (IRA)-tool.
In the following section the project partners Daimler and VW report on the application of HBMs as there are THUMS-D (Daimler) and THUMS-VW respectively which represent the main part of the virtual tool chain for pedestrian safety analysis. In some cases identical impact conditions were simulated and analysed such that a comparison of the results obtained by using similar HBMs under different crash codes was possible as well. However, a detailed comparison based on a selected load case is reported in D3.4. Numerous simulations were conducted by both partners which provided results in terms of kinematics, proposed HBM injury parameters and head injury risk assessment. When comparing the results, it was found that the injury risk for a subdural haematoma (SDH) predicted by THUMS-D was much lower than the SDH risk predictions by THUMS-VW. Other body regions however had comparable results, for example the maximum lateral rib deflections and pelvis forces. Larger differences were observed mainly at the higher impact velocity of 40 km/h and at the 20% offset pedestrian position for the tibia and thigh forces. Since completely identical HBMs could not be used due to the different crash software utilized differences were expected. As suggested in report D3.4 the variation of results should be reconsidered when identical THUMS models can be used in both crash codes in the future. However, the results obtained demonstrate that the entire tool chain is working properly. Each THUMS together with the SUFEHM allows a quantitative head injury risk assessment.
The analysis undertaken by CRF with the use of their Pedestrian Accident Compliant (PAC) FE dummy model in the virtual tool chain has been reported in detail as well. Similarly, results could also be compared with another model, this time in comparison to Daimler’s THUMS-D analysis on REVM1. When comparing the kinematics of the two pedestrian models, it became obvious that the PAC FE dummy shows different kinematics, mainly according to the head, which does not seem reasonable. Since the PAC dummy kinematics do not compare well with the kinematics derived from cadaver testing from the literature, the PAC FE dummy in its current version cannot be recommended to be used as part of the virtual tool chain for the assessment of pedestrian safety. A modification of the PAC FE dummy to overcome the issue of poor head kinematics was not planned and considered within this project.
When assessing pedestrian safety sensor systems in the virtual tool chain, in task T3.2 shortcomings regarding the analysed vehicle front structure of REVM1 become evident. The root cause was the weakness of the structural parts of the vehicle model. In consequence no virtual sensor evaluation was possible. Normally, state-of-the-art vehicle front ends have a stiffer structure due to e.g. strutting parts, radiator grill, etc.. This allows a better separation of “nofire” and “mustfire” signals in the virtual sensor model. An improved vehicle structure of REVM1 therefore would help as far as the application of the virtual sensor calibration approach is concerned. Hence, a virtual sensor calibration can still be a valid part of the virtual tool chain.
In addition, asphalt models have been introduced and partly used to better simulate the road rather than applying a simple rigid plane. However, no benefit could be seen in using these models when comparing head injury parameters on the HBMs. Nevertheless, the road modelling is important when the secondary impact is simulated and some kind of improved asphalt - in terms of energy absorption – will be introduced. This is planned to be further evaluated in work package WP4 of this project.
Summing up it can be concluded, that a further step is taken towards a virtual assessment of pedestrian safety using entire pedestrian models and especially HBMs. Currently, as far as specific injury risk assessments are concerned, this is limited to the head while using UNISTRA’s SUFEHM together with the post-processing IRA-tool. On the other hand, although currently being limited to the head with respect to a detailed and quantitative risk assessment, the most important body region as far as frequent severe or fatal injuries of pedestrians are concerned, can be assessed. For all other body regions further research is needed to get to model dependant injury risk assessments. The simple adaption of available injury parameter thresholds derived from tests found in the literature, e.g. a maximum plastic strain value for bone fracture, and their direct application to the post-processing of the HBMs, only allows a first rough estimation. It definitely cannot be a sufficient basis if the models are intended to be used within a virtual certification process yet. For this purpose widely accepted appropriate injury parameters and belonging model dependent thresholds do not yet exists, but are required and need to be agreed upon. Anyhow, the proposed injury parameters can be used for the optimisation in the virtual tool chain. The application of this virtual tool chain for pedestrian safety measures will be part of work package WP4 in this project.
In terms of the scattering of simulation results, the influence of small changes to the injury parameters and risk assessment could be shown by slight variation of the initial setups, i.e. position of the pedestrian relative to the vehicle. The scattering is either observed for a variation within the same model and code, but also for the same load case between the codes. Further robustness of the pedestrian HBMs as well as further improvement towards harmonised HBMs fulfilling a definite level of validation and being assessed by a kind of performance rating, is desirable. A first step to achieve a harmonised model and to define a specific validation catalogue is undertaken by the work of the THUMS User Community “TUC” where partners Daimler and VW are involved. Unfortunately, due to the timing, this first harmonised model could not yet be applied to the SafeEV project.
D3.6:
The aim of the work in WP3 of the SafeEV Project is to develop a seamless tool chain to investigate solutions for pedestrian safety and occupant protection in view of future virtual certification testing of Small Electric Vehicles (SEV). The work is based on the test conditions which were described in Deliverable 2 (D2) of the SafeEV Project. The work described in this report D3.6 concerns the work carried out in Task 3.3: Development of an advanced simulation methodology (tool chain) for a consistent safety analysis for occupant protection (incl. compatibility) in ‘Safe Electric Vehicles (SEVs). A sister report (D3.5) in SafeEV concerns a tool chain for pedestrian protection.
D3.6 is a follow-up of the report D3.3 where the objective was specifically to define requirements for such a tool chain. The main aim of the current report D3.6 is to report the work that was carried out in Task 3.3 based on these requirements in order to complete the tool-chain as well as to show some results (if applicable) to demonstrate the use of the tool-chain.
D3.6 presents tool-chains for the most important accident conditions for protection of car occupants: respectively frontal impacts, side impacts, front crash compatibility and side crash compatibility. Simulation of the occupant in the pre-crash phase as well a crash sensor modelling will become more and more important in a tool chain for the safety design of vehicles, due to the introduction of active safety systems. They will be described in separate Chapters in this report. Also post-processing is important in a tool-chain and will be described in a separate Chapter.
Both the tool-chains proposed for virtual assessment of frontal and side occupant safety respectively include a Human Body Model for the driver. The various simulations demonstrate that HBM occupant evaluation for SEV’s is possible including injury prediction.
The front and side crash compatibility research was carried out with crash dummy models: in fact, even if in the future applications HBMs are envisioned as the standard occupants within the virtual tool chain, the use of virtual dummies still will maintain in the future its relevance because of the link with the experimental crash testing. Then the inclusion of such more traditional virtual tools has been considered of importance for the development of the tool chain dedicated to SEVs.
Front crash compatibility assessment concerns questions of self and partner protection and the defined tool-chain was able to address both aspects. Also for side impact compatibility the suitability of the tool-chain was demonstrated. A compatibility index was proposed that showed a good capability to classify the level of compatibility of the SEV in both front and side compatibility. The overall behaviour of the evaluated SEV was in general adequate both for front and side impact compatibility and some possibilities to further improve the performance could be indicated.
A reliable tool-chain which involves also the pre-crash phase is important to study human response in the pre-crash phase. The objective was to study the tool-chain in an integrated pre-crash and in-crash situation. For the pre-crash phase realistic behaviour could be demonstrated in comparison with human volunteer testing but the continuation of the activated HBM in the in-crash phase caused instability problems in the HBM model (THUMS). So the current tool-chain has still limitations in the respect and therefore for the work in WP4 an alternative approach will be used.
The objective of the crash sensor modelling activity in SafeEV was the integration and assessment of frontal crash sensor system in the virtual tool-chain. The methodology could be successfully demonstrated in various crash conditions involving an SEV. For “traditional” vehicle architectures used in current small vehicles like the Smart fortwo architecture the current approach for sensor locations can be applied. In case of large changes of the vehicle architecture alternative sensor layouts have been suggested.
The pre-processing section of this report is focussing on 3 generic aspects of a crash simulation tool-chain: belt positioning, occupant positioning and simulations in a sled environment. The focus of the post-processing part is on how to deal with the vast amount of results for injury evaluation criteria produced by human body models (HBM), including the use of the Abbreviated Injury Scale (AIS).
D3.7:
The main objective of the work in WP3 of the SafeEV project is to develop a seamless tool chain to investigate solutions for pedestrian and occupant protection as a part of a virtual certification of Small Electric Vehicles (SEVs).
The work described in D3.7 concerns the work carried out in task 3.4: Revision of test configurations and evaluation criteria towards the new developed evaluation methods. This report is a follow-up of the report D2 and focuses on final test configurations (front, side and pedestrian impact scenarios) and evaluation criteria (occupant and vehicle), including the findings of WP3.
D3.7 presents a final proposal of test conditions and evaluations for virtual assessment of the most important accident conditions in urban environment related to SEVs in the future. Scenarios addressing occupant protection include front and side impacts and the compatibility of vehicles.
Having all the components of the evaluation tool chain, simulations were performed to identify the most appropriate test configuration. As the outcome, the final test configurations for the front and side impact test conditions are listed below.
The proposed frontal crash scenario, designed to asses occupant protection in SEVs, uses:
• Crash barrier: 1.300 kg pMPDB,
• Impact speed 35 km/h – 35 km/h,
• Impact angle: 30 deg
• Overlap: 50 %
• HBM in driver seat (Hybrid III or newer dummy model for comparison to experimental testing)
The following body parts are supposed to be observed:
• Head evaluation based on acceleration, stress, strain and energy
• Ribs, bones and ligaments fracture
• Heart, lungs, liver, spleen and intestines damage
The side crash scenario involves:
• Crash barrier: 1.300 kg AE-MDB
• Impact speed: 45 km/h
• Impact angle: 90 deg
• Impact point: R + 250 mm
• HBM in driver seat (WorldSID or newer dummy model for comparison to experimental testing)
The following body parts are supposed to be observed::Head evaluation based on acceleration, stress, strain and energy
• Ribs and pelvis fracture
• Ribcage compression
• Pelvic and abdomen forces
Despite the advancement of HBMs, virtual dummy models will remain in the testing protocols as a link to experimental testing.
Pedestrian protection is a vital part of the developed tool chain. As SEVs are intended to be used mostly in an urban environment, pedestrian protection evaluation is equally important as occupant protection. As a first step sensors which provide information to prevent accidents were evaluated. Secondly, in case an accident happens injury protection offered in and by the vehicle on a tissue level was analysed.
Based on this study, the final test configuration for virtual pedestrian protection is proposed as follows:
• Two velocities:
- 25 km/h to take pre-crash braking into account
- 40 km/h to simulate a non-detected pedestrian
• Impact locations with the H-point aligned to the relevant position:
- 20 % (or between wheel and fender for outstanding wheels)
- 50 % impact location of vehicle front
- 80 % (not used for outstanding wheels vehicle type)
- Directly on wheel for outstanding wheels vehicle type
- HBM pedestrian model
The following body parts are supposed to be observed:
• Head evaluation based on acceleration, stress, strain and energy
• Ribs and pelvis evaluation based on plastic strain
• Ribcage based on deflection of selected ribs
• Femur, tibia and fibula based on plastic strain
These configurations are described within this deliverable and will be used in Work Package 4 and 5 to assess the effectiveness of new generic solutions for occupant and pedestrian safety enhancement.
HBMs were used successfully in this work to numerically simulate kinematics and dynamics of occupant and pedestrian in various scenarios. Stress and strain values could be derived for various body parts. Nevertheless to fully benefit from the tool chain for a virtual certification specific injury risk curves and failure threshold values for relevant organs and individual HBMs need to be developed in future projects.
D4.1:
In D4.1 developed safety measures for reducing the injury risk for pedestrians during collisions with small electric vehicles are described. These safety measures are developed for two generic vehicle models, the so-called REVM1 and REVM2.
Following the virtual assessment approach developed in WP3, critical components for both vehicle models are defined in a first step. Afterwards solutions for each critical component are developed based on the virtual tool chain. Finally, the most promising concept, which will be further investigated in WP5, is defined and guidelines for pedestrian safety for small electric vehicles are derived.
Within the present report the potential of a stiffness adaptable CFRP is investigated. The investigated CFRP material has a thin thermoplastic coating between the carbon fibre reinforcement and the thermoset matrix resin. By applying an electric current through the carbon fibres, the thermoplastic coating is heated above the glass transition temperature, which results in reduced fibre/matrix load transfer and consequently in reduced stiffness. Due to this, the Young modulus in longitudinal direction can be reduced from 89 GPa in the stiff state to 22 GPa in the soft state.
For REVM1 safety measures for the bumper, windscreen and windscreen frame area are developed. Of special interests are the front bumper solution using a stiffness adaptable CFRP, a new approach for considering the stochastic fracture behaviour of glass and an external airbag for the windscreen frame area in an O-shape design. Furthermore, different sensor positions for improving the signal discrimination potential are considered. With the stiffness adaptable CFRP for the front bumper, the injury risks for the lower leg impactor could be drastically reduced (tibia acceleration from 302 g to 85 g, knee bending from 31 ° to 3.2 °, knee shearing from -7.1 mm to -2.0 mm). By considering the stochastic fracture behaviour of glass within the new assessment approach for laminated safety glass, the windscreen geometry could be improved and the resulting cumulated head injury criterion reduced from 805 to 577.
For REVM2 various safety measures for the lower and upper transversal beam, the lower windscreen cross beam and the outboard wheel design are developed. The usage of the stiffness adaptable CFRP is further investigated for the lower transversal and the lower windscreen cross beam. Also the effect of a foam covering the chassis of the outboard wheel design and the dependency of impact speed on injury risk are evaluated. With the stiffness adaptable CFRP used for the lower transversal beam the injury risks for the lower leg impactor could be drastically reduced (tibia acceleration from 263 g to 113 g, knee bending from 11 ° to 9 °, knee shearing from 3.5 mm to 3.8 mm). As the study will show, by integrating a foam the injury risk for lower leg fracture could be reduced until a velocity of 25 km/h. For higher impact velocities, the safety potential of foam paddings is rather low.
A further safety measure for reducing the injury risk during secondary impact on the ground is assessed based on two forgiving asphalt designs, developed by an external partner. The analysis showed that the usage of a full human body model for the assessment of injury risks during secondary impact is not feasible. Small changes during the primary impact strongly influenced the injury risk. The assessment using head models only showed a reduced injury risk for an impact on the forgiving asphalt designs.
Based on a comparison of the overall safety potential of the different measures, the lower transversal beam using stiffness adaptable CFRP showed the most promising potential, since it reduces the injury risk for a leg injury for pedestrians and might be beneficial for occupant safety during high speed collisions as well.
Finally, guidelines for pedestrian safety for small electric vehicles are derived.
D4.2:
D4.2 describes the results of SafeEV work package 4 subtask 2. The target of the subtask activities is the improvement of the occupant safety level for small electric vehicles demonstrated for the exemplary reference electric vehicle models (REVMs) available within the project.
Two types of improvements are discussed:
- Structural improvements reducing the passenger compartment intrusion level and enhancing the structural interaction capacities of the REVMs
- Optimisation of existing and the implementation of novel restraint system features with the aim to improve the interaction between the occupant and the vehicle structure in case of crash
The crash test configurations serving as design target for the vehicles are the ones defined and justified for SEV occupant protection testing within SafeEV work package 3. In addition, also further test configurations representing specific problems for the different REVMs are investigated. This results in a set of additional vehicle-to-vehicle crash configurations. In this way the discussed structural improvements can also be analysed with respect to their use potential in real world crash situations that are out of scope of laboratory testing.
The described improvements follow two targets. One is to augment the protection level for the occupants of small electric vehicles as represented by the given set of REVMs. The other is to contribute to the fulfilment of lightweight targets defined for electric vehicles where it is possible without detriment to the crash safety level.
For the improvement measures showing the best performance concerning the two targets further investigations are executed to describe additional questions such as their economic feasibility.
Having discussed all improvements for the REVMs, global design recommendations concerning the given crash test configurations summarise the activity results.
Future small electric vehicle design activities will be able to profit especially from these results in case the global structural design of their concepts is comparable to some structural aspects of the selection of REVMs. As the set of REVMs represents the state-of-the-art concerning global structural design parameters for small cars, the use benefits are expected to be high.
The structural improvement highlights at a glance:
• Aluminium space frame design front vehicle resulting in improved front crash safety and 40 % weight saving potential
• Advanced carbon fibre reinforced plastic (CFRP) material with modifiable stiffness properties applied to the front cross member resulting in balanced protection levels for occupants in case of front crash and for vulnerable road users in case of vehicle impacting
• Retractable steering wheel system allowing to enlarge the driver seat occupant survival space in front crash situations without weight-intensive structural reinforcements
• Active door beam with inflatable structure augmenting the occupant protection level in side crash situations without high design space demand
As result 14 global design recommendations are formulated concerning structural improvements to occupant safety and weight reduction for small electric vehicles.
The restraint system improvement highlights at a glance:
• Optimised driver and knee airbag behaviour increasing the safety benefit of a retractable steering wheel system for full width deformable barrier (FWDB) front crash
• Inflatable belt-bag reducing the risk of driver’s head and chest injuries in oblique mobile progressive deformable barrier (MPDB) front crash
• Pre-tensioning system for belts reducing the driver position change (out-of-position state) before frontal impact due to autonomous emergency brake system functionality
• Lateral movable seat equipped with an inflatable belt system reducing the driver’s head displacement and chest deflections and forces in side crash test configurations
Resulting from the executed investigations 8 global system design and design process recommendations are defined concerning restraint system improvements to the driver seat occupant safety.
D5.1:
D5.1 describes the results of SafeEV work package 5 subtask 1, in which a safety solution selected from WP4 is physically produced. A suitable test bench is then devised and used to evaluate the performance of the solution. As the selected component is a stiffness adaptable transversal beam for REVM2, testing is performed in both the reduced stiffness and the normal stiffness states for comparison. It is concluded that by reducing the stiffness of the material, a significant reduction in peak load transferred to the pedestrian can be achieved. Furthermore the original Crossbeam were also tested. Details of the development of the physical component and its test bench are described in this report.
Two virtual demonstrators are developed based on concepts previously discussed in WP4. These are an “O shaped” airbag that showed good potential for improving pedestrian safety, and the modelling of the restraint system configuration of REVM1.
An analysis of the REVM1 front structure updated in WP4 is performed to assess the influence of structural changes on the sensing system. A comparison is made between the design used in WP3, in which it was found the structure was too soft for state of the art sensor performance and the design from WP4. It is found that as well as increasing pedestrian safety (as shown in WP4.1) the time for activation of the pedestrian safety system is also improved.
D5.2:
D5.2 focuses on an improvement of the car front in order to increase pedestrian protection. Two versions of the car model were used to simulate a pedestrian accident in order to compare the car behaviour, the body kinematic and finally to assess comparatively the different injury risks posed. The proposed improvement is limited to the change of the front transversal beam. Therefore only lower leg injury risk was reduced. However, due to the slight changes of the body kinematic, increased head injury risk was computed. This illustrates the complexity of pedestrian protection optimization.
Even if the pedestrian protection was not significantly improved in this section, the subtask is a perfect demonstrator of the tool chain developed in framework of SafeEV.
D6.1:
D6.1 describes the intentions of the consortium members for disseminating and exploiting the project results and findings.
Within the dissemination plan strategy and planned activities are described. For these dissemination activities already possible audience and main addressees where identified within the implementation phase of the project. This report will strengthen now this information transfer in terms of a coordinated activity also within the SEAM cluster. Moreover feedback and comments on the proposed methods and concepts from relevant stakeholders and/or dedicated activities (e.g. workshops) in the course of the project will be taken into account for the formulation of the “best practice guideline” at the end of the project.
While this Dissemination Plan defines the planned activities and its coordination by the partners, the second part of this report presents a plan for using and implementing new knowledge and project findings in detail and related activities and how they support exploitation.
D6.2:
D6.2 describes the final workshop of SafeEV together with the SEAM cluster project MATISSE and attracted 46 participants. This deliverable provides details of when and how the workshop was organised and what the contents of the presentations were.
D8.1:
In the next 20 years the number of small and light-weight full electric vehicles will substantially increase especially in urban areas. These Small Electric Vehicles (SEVs) show distinctive design differences compared to the traditional car (e.g. no bonnets, vertical windscreens, outboard wheels). Thus the consequences of impacts of SEVs with vulnerable road users (VRU) and other (heavier) vehicles will be different from traditional collisions. These fundamental changes are not adequately addressed by current vehicle safety evaluation methods and regulations. Therefore it was the main objective of the project SafeEV, based on a systematic analysis and prognosis of future accident scenarios, to define advanced test scenarios and evaluation criteria for VRU, occupant safety and compatibility of SEVs. Moreover, industrial applicable methods for virtual testing of these scenarios and criteria (e.g. a method for active occupant safety assessment) were developed and demonstrated in terms of a virtual tool chain for potential extension of new normative scenario for this kind of vehicles.
Within D8.1 all project results are now reviewed and consolidated with the focus to formulate a best practice guideline for the implementation of the advanced simulation methodologies into such a “tool chain” for enhanced safety analysis concerning pedestrian and occupant safety on the one hand and also global design recommendations for dedicated and advanced safety solutions for SEVs (derived and evaluated through the application of the projected simulation methodologies) on the other hand.
Notwithstanding the main objective and above mentioned focus on virtual methods of the project the report follows now the “function chain” of an integrated safety system (traffic / test scenario > pre-crash sensing & safety system > in crash safety system). For a certain function level then the components of the virtual “tool chain” are discussed.
Best practice guide for the use of stochastic analysis and simulation to derive and predict future accident and finally test scenarios are summarized in the first part of the report. Lessons learned are discussed in terms of linking these results with other scenario and prediction methodologies.
It could be demonstrated, that virtual methods facilitated the optimization of environmental sensing in terms of scenario definition (pre-crash) and are indispensable for in-crash optimization of a crash sensor cluster (pedestrian & occupant safety) concerning interaction with vehicle and structural parts. Finally it could be stated that pre-crash sensor based autonomous braking systems should be mandatory for SEVs respectively vehicles with dedicated use in urban areas.
Structural interaction and compatibility concerning ideal load path are identified with virtual models and general design guideline are given. For this extensively the virtual tool chain including V & V (Validation & Verification) aspects are described. Especially the issue of formulating novel material models are addressed by the development of advanced and active structural components.
Safety and protection systems for pedestrians and occupants are discussed on the next functional level. In general it could be stated that any distribution of load (e.g. kneebag, inflatable belt) within the crash phase is beneficial in terms of best possible restraint effect. This also applies for advanced functions on the temporal axis. It could be demonstrated that pre-crash triggered restraint and / or pre-acceleration followed finally also by adaptive load limiting (e.g. retractable steering wheel) are effective strategies for improving the protection level offered by SEVs. Nevertheless, especially the use of virtual methods has shown, that combining such advanced systems an intelligent system management will be needed.
Beside a more standardised evaluation (a post processing tool and method was proposed in the course of the project) the projected numerical tools and models offered even more detailed information whenever specific effects were assessed or method for system parameter optimization was needed. On the other hand, it looked like, that the applied Human Body Models were quite sensitive to system parameters or vehicle design (this was observed especially for the pedestrian safety evaluation). Finally it has to be stated, that for most body regions of the applied HBM no general criteria are defined so far. However, it was possible to work effectively with the proposed injury parameters.
In the last chapter also general recommendations (e.g. comparability of simulations codes, validation of virtual models) and needs for further research are given.
Based on the results of this project, it is entirely realistic that virtual methods and tools could be implemented in evaluation and assessment procedures for specific SEVs respectively for a new class of vehicles.
Potential Impact:
Beside the common dissemination activities and in support to the communication activities of the Commission services the consortium may be requested (on an annual basis or upon request) to provide the Commission with a 2 pages information sheet which will be drafted in a standard format communicated by the Commission. The Commission services may also request on illustration (picture, schema or drawing) to illustrate such communication material.
A first project information was requested by the Commission in October 2013 for the edition of the SST projects synopsis, which was responded accordingly. The project is also asked to contribute to relevant dissemination, international cooperation and information exchange activities of interest to the Commission such as those within the scope of the European Green Cars Initiative. Also for this EGVI dissemination area already a contribution was prepared. SafeEV has provided a slideshow which overviews concisely their objectives and major outcomes to the European Green Vehicles Initiative Association for the TRA 2014 in Paris. The key audience and main addressees of dissemination activities are already defined in the Description of Work respectively during the implementation phase of this project. Nevertheless, this list will be updated continuously in the course of the project and will follow also feedback from dedicated and cooperative activities (SEAM workshops, network activities etc.).
Most of the activities started in month one. Five main categories were identified to build the core of the dissemination strategy of this project.
Project oriented dissemination & information platform
Just after launch the project SafeEV had its own website (www.project-safeev.eu) making project information available online. It was further developed during the first project period and will be regularly updated throughout the project. The site is interactive and contains project and partner description. It also builds the “first point of contact” for any interested parties which intend to get in contact with the consortium or searching for results, information to download or further links.
For this a newsletter was created, which will be send regularly to registered recipients. The registration can be done via this website.
Also all public reports listed on the website and will be made available by direct download. Finally also direct links to the SEAM cluster and the individual SEAM projects are established on the website.
Scientific publications and contribution to conferences
The participation in international conferences and contribution to scientific publications makes it possible to communicate SafeEV objectives and findings to wide expert public. The academic partners will especially contribute and prepare publications to conferences and journals. Target conferences are e.g. SAE, ESV. Most of the (scientific) partners in the consortium are closely and well involved in the scientific community or even members of a programme committees or editors of mentioned journals which will make it possible to submit a number of specific publications and issues on project findings and results.
This list follows the dissemination strategy and further road map for publications below:
First year dissemination
- Generate attention for SafeEV to wide expert public
- Communication of project objectives and tools & methods which will be used
Second year dissemination
- Communication of first project results (scenarios / prognosis → basis for further activities).
- Demand information & feedback from expert public and related initiatives / research
Third year dissemination
- Ongoing status report / results from project core activities
- Communication of first project findings & recommendations
Final / Fourth year dissemination
- Final project results and recommendations
- Provide guidelines to expert public and other stakeholders
- Initiate follow up activities (including SEAM Cluster)
- Initiate exploitation of results (including SEAM Cluster)
Implementation of project findings – interest groups & relevant stakeholders
Mainly the industrial partners (Daimler, CRF, VW, Pininfarina, Bosch) will initiate and coordinate cooperative activities in the project phase when it comes to implementation, demonstration and exploitation of the project results. This will be done via communication and direct transfer of project findings to interest and working groups (e.g. TUC, ACEA, VDA, CLEPA). Also relevant stakeholders in the area of vehicle and traffic safety (e.g. NCAP, EEVC, UNECE) will be addressed via dedicated activities.
Similar to the communication via the scientific community (3.2) the current network activities of the (industry) project partners will be used as a first level dissemination channel.
SafeEV activities are tracked in the TUC – THUMS USER COMMUNITY (www.tucproject.org). Focus is especially on the discussion and definition of dedicated methods for post processing of HBM simulation results and injury risk assessment. TUC represents a group of THUMS user, namely Audi, Autoliv, BMW, Daimler, Opel, Porsche, Toyota, VW and LMU (Ludwig Maximilians University Munich) as coordinator, which initiated this platform (Collaborative Project) end of 2012 to set up a framework and harmonize general and administrative requirements for the implementation of FE- Human Body Models in vehicle and traffic safety applications.
SafeEV activities are also communicated to and within PDB (www.pdb-org.com) - Partnership for Dummy Technology and Biomechanics, which was founded in 2002 as a cooperative venture between German car manufacturers in the competitively neutral field of crash test dummy technology, biomechanics and simulation. Since Feb. 2014 a sub-group addressing the use and general implementation of virtual HBM to development and assessment methods is installed in this organisation.
A direct link via a project partner also exists to the EEVC –European Enhanced Vehicle-Safety Committee and its WG 12 “Crash Dummies”. Another relevant EEVC WG in terms of exploitation of project findings is WG 22 “Virtual Testing”.
Beside these direct transfer of project results by personal / partners contact workshop events are seen as most appropriate activity.
The exploitation strategy of this initiative is mainly determined by the specific character and content of the research and project objectives.
On the one hand the main objective of this project is to develop and demonstrate a complete new assessment methodology for a vehicle class (or mobility mode), which up until now are not subject of regulatory or consumer testing requirements. So, ideally, stakeholders in this area might take offer the complete approach respectively methodology, tools and criteria (“tool chain”).
For this, this core result could be seen and defined as a pilot for a new generation and widely accepted assessment method. The basis for such a pilot is already provided by a long term process for which IMVITER delivered latest and relevant requirements. This general process for implementation of Virtual Testing in safety regulations and/or assessment was initiated 2006 by the CARS21 initiative and continuously further developed within research projects and even the EC regulatory framework.
Figure 4.1 illustrates this background and the optional continuation by a “SafeEV pilot”.
Figure: Steps in the implementation of VT in safety regulations and – assessment and context of a “SafeEV pilot”.
On the other hand, SaveEV is also providing all relevant components respectively “chain links” for a virtual testing environment & tool chain needed for the evaluation of pedestrian & occupant protection and increased compatibility of SEVs. So, also exploitation of partial results and specific findings will be promoted by project activities and even more by exploitation plans of the partners. The next paragraph highlights especially these more specific areas and dedicated project findings.
List of Websites:
http://www.project-safeev.eu/