SECOND CYCLE, 30 CREDITS STOCKHOLM SWEDEN 2019,
Development of a new test methodology for
car-to-truck crash
MATAS BUZYS SARA NILSSON
KTH ROYAL INSTITUTE OF TECHNOLOGY
SCHOOL OF INDUSTRIAL ENGINEERING AND MANAGEMENT
Development of a new test methodology for car-to-truck crash
Matas Buzys Sara Nilsson
Master of Science Thesis TRITA-ITM-EX 2019:246 KTH Industrial Engineering and Management
Machine Design SE-100 44 STOCKHOLM
ii
iii
Examensarbete TRITA-ITM-EX 2019:246
Utveckling av ny test metodik för bil mot lastbil kollision
Matas Buzys Sara Nilsson
Godkänt
2019-06-11
Examinator
Ulf Sellgren
Handledare
Ulf Sellgren
Uppdragsgivare
Scania
Kontaktperson
Johanna Åstrand
Sammanfattning
Till följ av de stora skadorna som riskeras vid frontalkollision mellan personbil och lastbil, utför Scania CV AB kraschtester för att bättre kunna utveckla komponenter med syfte att skydda passagerarna i personbilen. Den typ av test som denna studie bygger på utvärderar den s.k.
FUP:en (engelska Front Underrun Protection). I dagsläget görs ett fullskaligt test, där en personbil avfyras in i en lastbil. Syftet med studien är att undersöka möjligheten att utveckla en förenklad test metod där endast de väsentliga komponenterna från lastbilen inkluderas, och en representativ struktur ersätter personbilen. Om möjligt kommer detta minska kostnaderna samt möjliggöra för större repeterbarhet. Tester och utvärderingar görs med hjälp av simulationer i LS-Dyna, ANSA & META, och designkoncept visualiseras i CAD-programmet CATIA V5.
Resultat visar att det finns goda förutsättningar för att ersätta personbilen med en barriär av honeycomb struktur samt att lastbilen kan ersättas med en vagn där de väsentliga komponenterna fäst. Diskussioner kring simuleringarna och designen lyfter fram faktorer som visar på goda utvecklingsmöjligheter, men med betoning på det fortsatta arbetet som krävs.
Nyckelord: FUP, Frontalkrock, Honeycomb struktur, Kraschtest
iv
v
Master of Science Thesis TRITA-ITM-EX 2019:246
Development of new test methodology for car-to-truck crash
Matas Buzys Sara Nilsson
Approved
2019-06-11
Examiner
Ulf Sellgren
Supervisor
Ulf Sellgren
Commissioner
Scania
Contact person
Johanna Åstrand
Abstract
Scania CV AB are developing components to prevent fatal damages during frontal collisions with passenger cars. Therefore, they need to test their assemblies and specifically FUP (Frontal Underrun Protection). Currently, a full-scale test is done in which a passenger car is launched into a truck. The purpose of this study is to examine and develop the possibility of having a simplified test procedure in which only the relevant components of the truck are included, and a representative structure replaces the car. If possible, this would reduce costs and allow for greater repeatability. Analysis and evaluations are done via finite element models using ANSA, LS-Dyna and META. The conceptual design is visualized using CATIA V5. Results show good indication that the passenger car can be replaced by a trolley with deformable barriers mounted on it and the truck can be replaced by a simplified structure with main FUP components mounted onto it. Discussions about the numerical models results and the conceptual design highlight factors that show promising possibilities, but with emphasis on the continued work that is required.
Keywords: Crash tests, FUP, Frontal impact, Honeycomb structure
vi
vii
FOREWORD
We are grateful to our supervisor from KTH Royal Institute of Technology Ulf Sellgren who guided us through the thesis. We would also like to express our gratitude to Johanna Åstrand, Dan Loftén and Mikael Littmann who provided us with all the information needed to reach our goals and for advice regarding industry standard procedures. Furthermore, Jakob Leygraf for recommendations regarding bolt calculations and Caroline Svedberg for help with CATIA.
Additionally, all other people from Scania CV AB and TASS International who have contributed. Finally, we would like to thank Dominykas Gudavičius for support and guidance with FEM tools, such as ANSA and LS-Dyna.
Matas Buzys and Sara Nilsson 27th of May 2019
viii
ix
NOMENCLATURE
Notations
Symbol Description
Abolt Area of bolt (mm2)
α Lead angle of thread (degrees)
β Thread angle (degrees)
D Average diameter of nut’s contact area with the foundation (mm)
Dh Hole diameter (mm)
dm Average diameter of bolt (mm)
e Restitution coefficient
Ek Kinetic energy (J)
ε Plastic strain
F Force (N)
Fa Axial force from M (N)
Fimp Impact force (N)
k Spring coefficient (N/m)
µ Friction coefficient
µg Friction coefficient of thread when β=0 µg’ Friction coefficient of thread when β≠0 µu Friction coefficient of nut
m Mass (kg)
M Tightening torque (Nm)
N Key width (mm)
n Number of bolts
r Radius (mm)
σ Von Mises criterion stress (MPa)
σb Tensile stress (MPa)
σs Shear strength (MPa)
v Velocity (m/s)
U Strain energy (J)
x Deformation (mm)
x
Abbreviations
ADAC Allgemeiner Deutscher Automobile-Club (General German Automobile Club)
AE-MDB Advanced European Mobile Deformable Barrier
CAD Computer Aided Design
CAE Computer Aided Engineering
DOF Degrees of Freedom
EECV European Enhanced Vehicle-safety Committee Euro-NCAP European New Car Assessment Program
FEA Finite Element Analysis
FEM Finite Element Model
FUP Front Underrun Protection
IIHS Insurance Institute for Highway Safety
MDB Mobile Deformable Barrier
MPDB Mobile Progressive Deformable Barrier NHTSA New car Assessment Program
O-MDB Oblique Mobile Deformable Barrier
PID Property Identity
SDOF Single Degree of Freedom
xi
xii
TABLE OF CONTENTS
Sammanfattning ... iii
Abstract ... v
FOREWORD ... vii
NOMENCLATURE ... ix
1 INTRODUCTION ... 1
1.1 Background ... 1
1.2 Purpose ... 2
1.3 Research questions ... 2
1.4 Delimitations ... 2
1.5 Method ... 3
2 FRAMEOFREFERENCE ... 5
2.1 Types of test setup ... 5
2.2 Energy absorbing structures ... 6
2.2.1 Types ... 6
2.3 Deformable barriers and their variants ... 3
2.3.1 Types of barriers... 3
2.3.2 Mobile Progressive Deformable Barrier ... 3
2.3.3 Advanced European Mobile Deformable Barrier ... 6
2.3.4 IIHS Side Impact Barrier ... 8
2.3.5 Oblique Mobile Deformable Barrier ... 9
3 IMPLEMENTATION ... 10
3.1 Requirements ... 10
3.2 Concept generation and evaluation ... 10
3.2.1 Concepts ... 10
3.2.2 Evaluation ... 11
3.2.3 Selection ... 19
3.2.4 Test-rig setup ... 19
3.3 Selection of barrier ... 20
3.3.1 AE-MDB ... 20
3.3.2 MPDB ... 20
3.3.3 IIHS Barrier & O-MDB ... 20
3.3.4 Designing new and customized barrier ... 21
xiii
3.3.5 Selection of barrier and possible modifications ... 21
3.4 Truck structure: existing trolley ... 22
3.4.1 Limitations and requirements of the trolley ... 22
3.4.2 Bus frame from TASS ... 22
3.4.3 Mounting solution ... 23
3.5 Truck structure: customised trolley ... 25
3.5.1 Limitations and requirements of the trolley ... 25
3.5.2 Design of trolley ... 25
3.6 Selection of truck components ... 26
3.6.1 Components of interest in terms of deformation ... 27
3.6.2 First level of additional components ... 27
3.6.3 Second level of additional components ... 28
3.7 Numerical analysis setup ... 29
3.7.1 Materials ... 29
3.7.2 Geometry... 30
4 RESULTS ... 34
4.1 Simulation ... 34
4.1.1 65 km/h ... 34
4.1.2 70 km/h ... 2
4.1.3 70 km/h – 2nd geometrical setup ... 3
4.1.4 75 km/h ... 2
4.2 Truck structure ... 2
4.2.1 Bus frame ... 2
4.2.2 Customised trolley ... 4
4.2.3 Customised trolley –simulations ... 7
5 DISCUSSIONANDCONCLUSIONS ... 10
5.1 Discussion ... 10
5.1.1 Simulation ... 10
5.1.2 Mounting solution ... 10
5.2 Conclusions ... 11
6 RECOMMENDATIONSANDFUTUREWORK ... 12
6.1 Recommendations ... 12
6.2 Future work ... 12
7 REFERENCES ... 13
xiv
xv
FIGURES
Figure 1: Example of the dangers of frontal crash without underrun protection [3]. ... 1
Figure 2: Illustration of the method and its process steps. ... 3
Figure 3: Solutions of test setup; 1. Sled on track [4], 2. Pendulum [5], 3. Hydraulic actuator [6] and 4. Barrier mounted on trolley [7]. ... 5
Figure 4: Honeycomb load vs deformation capacity [8]. ... 6
Figure 5: Honeycomb sandwich structure [3]. ... 6
Figure 6: Cell configuration [8]. ... 7
Figure 7: Crush strength vs impact velocity relationship [8]. ... 8
Figure 8: Crush strength vs angle of applied load [8]. ... 8
Figure 9: Crushed aluminium honeycomb core [3]. ... 8
Figure 10: The three regions of the stress-strain curve. ... 2
Figure 11: View of MPDB and trolley to be implemented in the Euro-NCAP protocol [16]. .. 4
Figure 12: Final MPDB used in the ISUZU study [17]. ... 5
Figure 13: The honeycomb structure of the ODB [20], left, and MPBD [21], right. ... 5
Figure 14: The ODB and MBD [22] in usage, left and right respectively. ... 5
Figure 15: Dimension of blocks of the MPDB, width of barrier: 1000 mm [31]. ... 6
Figure 16: The fourth version of the AE-MDB, called AE-MDB v3.9 [23]. ... 7
Figure 17: Dimensions of each block of the AE-MDB [18]. ... 7
Figure 18: IIHS barrier to the left [26] and FMVSS 214 barrier to the right [27]. ... 8
Figure 19:IIHS Barrier with trolley [34]. ... 8
Figure 20: Oblique Mobile Deformable Barrier, OMDB, with trolley (image provided by Plascore). ... 9
Figure 21: Properties of the OMDB (image provided by Cellbond). ... 9
Figure 22: First concept. ... 11
Figure 23: Second concept. ... 11
Figure 24: UPN-80 beam cross-section [19]. ... 12
Figure 25: Energy distribution after impact. ... 14
Figure 26: Energy ratio vs time. ... 14
Figure 27: Force vs displacement. ... 15
Figure 28: Force vs time. ... 15
Figure 29: Force vs displacement. Comparison between two models. ... 16
Figure 30: Force vs displacement (velocity impact). ... 17
Figure 31: Force vs time (velocity impact). ... 17
Figure 32: Force vs displacement (mass variation). ... 18
Figure 33: Force vs time (mass variation). ... 18
Figure 34: Setup of full-scale vehicle test (image provided by Scania). ... 22
Figure 35: Bus frame proposed by TASS to be used as truck trolley (image provided by TASS). ... 23
Figure 36: Drawing of top of the bus frame (image provided by TASS). ... 23
Figure 37: Details of the different widths along the chassis beams. ... 23
Figure 38: Simplified illustration of mounting solution... 24
Figure 39: The initial set of components from which the 2 levels were constructed. ... 27
Figure 40: Parts of interest in crash test. ... 27
Figure 41: 1st level of setup of components. ... 28
Figure 42: 2nd level of setup f components. ... 28
Figure 43: Von Mises yield criterion in planar view [23]. ... 29
Figure 44: Stress-srain curve. ... 30
Figure 45: 1st setup top view. ... 31
xvi
Figure 46: 1st setup isometric view. ... 31
Figure 47: 1st setup side view. ... 31
Figure 48: 2nd setup top view. ... 32
Figure 49: 2nd setup isometric view. ... 32
Figure 50: 2nd setup side view. ... 32
Figure 51: Energy distribution (65 km/h). ... 34
Figure 52: Force-time relationship. ... 35
Figure 53: Force-displacement relationship. ... 35
Figure 54: Average force plot. ... 36
Figure 55: Displacement of barrier during impact. ... 36
Figure 56: Velocity of a moving trolley from the point of impact. ... 37
Figure 57: Acceleration of a moving trolley from the point of impact. ... 37
Figure 58: 3D deformations (65 km/h). ... 37
Figure 59. Energy distribution (70 km/h) ... 2
Figure 60: Force-displacement relationship. ... 3
Figure 61: Force-time relationship. ... 3
Figure 62: Average force. ... 4
Figure 63: Displacement of barrier during impact. ... 4
Figure 64: Velocity of moving trolley from the point of impact. ... 5
Figure 65: Acceleration of a moving trolley from a point of impact. ... 5
Figure 66: 3D deformations (70 km/h). ... 5
Figure 67: Impacting trolley acceleration. ... 2
Figure 68: Energy distribution (70 km/h) – different position. ... 3
Figure 69: Force-displacement relationship. ... 3
Figure 70: Force-time relationship. ... 4
Figure 71: Average force. ... 4
Figure 72: Displacement of a barrier during impact. ... 5
Figure 73: Velocity of a moving trolley from a point of impact. ... 5
Figure 74: Acceleration of a moving trolley from a point of impact. ... 5
Figure 75: 3D deformations (70 km/h). ... 2
Figure 76: Energy distribution (75 km/h). ... 2
Figure 77: Force-displacement relationship. ... 3
Figure 78: Force-time relationship. ... 3
Figure 79: Average force. ... 4
Figure 80: Displacement of a barrier during impact. ... 4
Figure 81: Velocity of a moving trolley from a point of impact. ... 5
Figure 82: Acceleration of a moving trolley from a point of impact. ... 5
Figure 83: 3D deformations (75 km/h). ... 5
Figure 84: Simplified visualisation of cross beams, detail of possible bolt hole on bottom flange of chassis beam. ... 3
Figure 85: Frame of truck trolley. ... 4
Figure 86: Front axis' spring link's rear connection. Bracket and connection are circled. ... 4
Figure 87: Connection to air bellows, detail of the mounting to the right. ... 5
Figure 88: Connection of front axis' spring link. ... 5
Figure 89: Connection of front of rear spring link. Arrows point to bolt holes which connects the bracket to the trolley frame, left and right. ... 5
Figure 90: Mounting of chassis beams to trolley. Arrows indicate bolting via bottom flange and to back plate. ... 6
Figure 91: Truck trolley, viewed from rear, with space for instruments highlighted. ... 6
Figure 92: Full truck trolley with all parts included. ... 7
Figure 93: Translational displacement of trolley due to loading. ... 7
Figure 94: Stresses of the trolley. Maximum value of 84 MPa. Front structure. ... 8
xvii
Figure 95:Stresses of the trolley. Maximum value of 84 MPa. Side structure. ... 8
Figure 96: Load frame implemented in current test setup. ... 14
Figure 97: Load frame and weights ("cigars" and blue TASS weight). ... 14
Figure 98: Remote control of braking system. ... 14
Figure 99: Sensor amplifier placed in the front of truck during current test setup. ... 14
xviii
TABLES
Table 1. Requirement specification. ... 10
Table 2. Beam measurements [12]. ... 11
Table 3. Input values. ... 13
Table 4. Maximum displacement at roughest conditions (>60 km/h, >10000 kg) ... 16
Table 5. Main parameters of interest. ... 29
Table 6. Acceleration comparison with values from a live test. ... 2
Table 7: Results of bolt dimensioning. ... 2
Table 8: Overview of number of cross beams required in relation to bolt size. ... 3
1
1 INTRODUCTION
This chapter presents the background, purpose and methodology of the thesis. For additional information regarding the planning and risk analysis, see Appendix A and Appendix B.
1.1 Background
Scania CV AB is a Swedish manufacturer and supplier of trucks and buses, with sales and provided services in over 100 countries. Together with MAN and Volkswagen Caminhões e Ônibus, Scania constitutes the umbrella company TRATON, which is a subsidiary of Volkswagen group. TRATON aims to be a global champion of the industry for the next decade and to achieve this, they have taken up a strategy to have all the heavy vehicle brands bundled under one roof to speed up the decision making. Moreover, they aim to have jointly developed products for Scania and MAN. For example engines, exhaust systems, transmissions etc. In addition, Scania aims to be an innovator within the market regarding safety measures.
Due to the great mass in road accidents involving heavier vehicles such as trucks and busses, the injuries are usually much more severe than with accidents involving only lighter vehicles, e.g. passenger cars. In 2015, 14 % of road fatalities in the EU involved heavy goods vehicles (of which approximately 3 % were represented by buses or coaches). Of those who died in accidents involving heavy duty vehicles, 50 % were traveling by car [1]. In accidents between a truck and a car, the majority occur as a frontal crash.
Since 2003 all new trucks in European Union must have installed Front Underrun Protection (FUP). It is an energy absorbing structure, which also acts as a tool to prevent passenger cars from going under the truck during a frontal collision. The regulations within ECE 93 [2] state that the structure is allowed to deflect up to 400 mm in rearward direction. In Figure 1 the possible dangers of when a FUP isn’t included is demonstrated.
In order to ensure the function of the FUP full-scale crash tests are performed. These involve entire vehicles which are used to re-enact the frontal impact of a passenger car on a truck. Due to the large setup of the crash test and the usage of full-scale vehicles, the tests become expensive and therefore less repeatable. They also become time consuming due to preparations and the fact that the test methodology has to be kept the same as the previous one for test validity. Furthermore, due to the cost of the tests, they are performed rarely.
Figure 1: Example of the dangers of frontal crash without underrun protection [3].
2
1.2 Purpose
The purpose of this thesis is to examine the current test method, identify the essential components and based on this design a test rig. The essential components should include those of interest (tested in the original setup) as well as those required to obtain consistent results with the full-scale test. The aim of the test rig is to allow for tests with only the relevant components, eliminating the need for complete vehicles. As a result, costs are reduced, repeatability is increased and the way of testing becomes more diverse. The test focusing on a few components will enable more and faster tests during development, while the full-scale test will still be used for verification of the whole system and not singular components.
The deliverables are to:
● Study the current methods at Scania.
● Develop a method for rig test.
● Design the test rig:
o Design of setup.
o Design of truck structure.
o Design of car structure.
● Verify the test rig with simulation and compare results with a full-scale test, if possible.
1.3 Research questions
● Is it possible to develop a cost efficient test device with repeatability possibilities?
● Is it enough to test only specific parts of the truck?
● What is the accuracy of the test-rig?
● Can a deformable barrier accurately represent real world conditions in car to truck crash test?
● How can simplified structures be incorporated in the car to truck crash test?
● Can honeycomb structure represent the front of a car during a crash test?
1.4 Delimitations
● Underrun protection is the main focus and should be the priority in components to be tested in the new test method. Any additional request should only be included if possible to do so without interfering with this and the other delimitations.
● No need for full chassis. The aim is as stated to create a test setup using as few components as possible.
● Conceptual design of truck structure. The design should be enough to evaluate if it has potential for further development, but no detailed design is required.
● No crash dummies are required. The values of the accelerometer will be used to determine if they are within standard regulations.
● Simplified design.
● FEM analysis of barriers acting as impactors.
3
1.5 Method
The method consists of nine steps. The following paragraph briefly explains each step of the process and their respective purposes. For an illustration of the overall method with process steps, see Figure 2. The first step includes gathering of information, in which interviews are held with relevant people at Scania and data of current and previous tests are collected. Using the information obtained in the first step, focus points are identified. For example determination of what components are required to be included in the rig. This helps keep the project within the specified scope. In the third step, research alternatives are investigated. Using relevant articles and additional sources, current and previous ideas and solutions are reviewed. In the final step before concept generation, the collected information from Scania and the relevant research obtained are compared and evaluated. The concept generation is constructed based on the knowledge gathered in the previous steps. The concepts are evaluated and when a final concept is approved, the design of the rig is initiated. Simulations are made to check if methodology is viable. The design is revised until the results are consistent.
In order to perform the simulation and design of the test rig, certain software is used. These include CATIA, ANSA and LS-Dyna. Any additional knowledge required to be obtained concerning these will be done alongside information gathering.
CATIA is a software with several possible implication areas. In this thesis it is used as a computer aided design tool, CAD. It enables for a 3D model of a component or an assembly of components to be designed, viewed and/or evaluation via e.g. simulations. The software can be used to both create a more conceptual design, to visualise an idea or perform a rough evaluation, or a detailed design with drawings ready to send for manufacturing. As mentioned previously, this thesis focuses on a more conceptual design than detailed. During the project 3D-parts of Scania will be available in order to take into consideration the possible implementation of these in the design.
ANSA, LS-Dyna and META are software used for FEM analysis. As within the order of listing the first one is a pre-processor. It is a program that creates an output from a certain input, which then could be used by another program. Simply put, ANSA is used as visual tool to prepare finite element model for the solver by meshing, setting up boundaries and load conditions and other physical necessities to achieve the desired outcome. LS-Dyna inputs data prepared by ANSA and calculates the output. This software has multiple capabilities, but in this thesis it is only used as a solver. Finally, META is a post-processor tool, which helps to interpret numerical values from LS-Dyna and provides means to create visual output and all the needed graphs.
Figure 2: Illustration of the method and its process steps.
4
5
2 FRAME OF REFERENCE
In this chapter the background study is presented. It focuses on current test setups for crash testing with trucks and passenger cars.
2.1 Types of test setup
In order to initiate the concept generation different types of test setups implemented in the industry have been examined. This section introduces four possible ones to be used as base for design of the new test method. Figure 3 illustrates four examples of setups.
1. Sled on track test provides motion within a desired direction and provides a controlled way of movement. Parts that need to be tested can be mounted on a carriage, which is accelerated to a desired velocity and crashed into a structure.
2. Pendulum test rig is used to check deformation of a structure under a certain impact load. It is a large structure which requires to be able to hold heavy masses and reproduce rotational movement. The impacting barrier, with a heavy mass is released from a predefined position and crashed into an object of interest.
3. Hydraulic actuator is used as a stress test. It is mostly used to check if the structure is capable to withstand a certain load.
4. Barrier mounted on a trolley represents a standard passenger vehicle. The barrier structure is used to represent the front of a typical car and tries to replicate its mechanical properties, for example stiffness as close as possible. The trolley is accelerated and crashed into a structure.
Figure 3: Solutions of test setup; 1. Sled on track [4], 2. Pendulum [5], 3. Hydraulic actuator [6] and 4. Barrier mounted on trolley [7].
The first one, sled on tracks, will not be able to represent the structure in a desired form, because the motion is allowed in only one direction. Thus, it will not represent an actual car-to-truck impact and majority of the impact forces might be absorbed by the sled structure. The second structure, pendulum, requires a sufficient amount of space, where it can be built. Additionally, the pendulum is not able to represent accelerations acting within the vehicle during a crash.
Furthermore, if there is a desire to expand the test and use crash dummies, another test-rig must
6
be used. Hydraulic actuator setup cannot replicate the impact. It is capable to create continuously increasing load, however, it lacks accelerations which are acting during a crash and therefore strain rates will not be the same. Since the barrier is the best alternative to represent mechanical properties of a car’s front, a more detailed investigation of this is continued.
2.2 Energy absorbing structures
There are several types of energy absorbing structures made of metallic or non-metallic materials. Honeycomb is the preferred one due to properties such as uniform crushing under a specific load, highest crush strength-to-weight ratio and a long stroke. They act as nearly perfectly plastic structures. Figure 4 shows the relationship between load and deformation of such an absorber, where the area below the curve represents absorbed energy.
Figure 4: Honeycomb load vs deformation capacity [8].
Bottoming out occurs when the object is not able to absorb the whole initial kinetic energy or absorption capacity is exceeded.
2.2.1 Types
Most common one used for vehicle crash testing are sandwich structures of aluminium honeycomb core due to the relatively high quality-cost efficient properties. An example is shown in Figure 5. Aluminium core usually has a crash stroke between 70 – 80% of its height [8]. These structures often differ by crush strength, manufacturing process and cell combination. Regularly these properties correlate.
Figure 5: Honeycomb sandwich structure [3].
7 Manufacturing process:
Adhesive bonding
Resistance welding
Brazing
Diffusion bonding
Thermal fusion
Most popular process of manufacturing is adhesive bonding, because it is the cheapest and easiest solution when compared to other ones. However, welding, brazing and others are employed when temperature plays a critical role in honeycombs function.
Cell configurations
Hexagon is the most common type manufactured using adhesive bonding. Square shapes are either brazed or resistance welded. Flex-core is used when honeycomb needs to be wrapped around spherical shapes. Over-expanded is a regular hexagon structure expanded into a square one. Please see Figure 6 for graphical reference.
Figure 6: Cell configuration [8].
Crush strength
As can be seen in the Figure 7 crush strength depends on the impact velocity, when the latter one increases the other one does as well. This graph shows specific honeycomb crush strength – velocity relationship. Other honeycomb structures might act differently due to variation in material properties, such as density, etc. At high impact velocities dynamic crush strength is about 30% higher than the static [8]. However, the crush strength drops if the load is not applied parallel to the cells. Figure 8 shows how an angle application affects the crush strength.
8
Figure 7: Crush strength vs impact velocity relationship [8].
Figure 8: Crush strength vs angle of applied load [8].
Figure 9: Crushed aluminium honeycomb core [3].
Properties:
Thickness and crush strength are inversely proportional. Meaning; when thickness increases crush strength decreases.
Thickness does not have a huge influence on compressibility or shear moduli.
Hexagonal cell honeycomb is orthotropic.
Honeycomb is neither homogeneous nor isotropic.
Honeycomb strength provided within product brochures is always ultimate failure stress.
2 Factors
Behaviour of sandwich structures under loading depends not only on the core and skins, but also on the adhesive properties between the core-skin bonds. Impact velocity has a remarkable influence on the structure. Therefore, assumptions taken from low velocity impact cannot be used when high velocity crash is present. For example, high velocity impacts are independent of boundary conditions, but vice versa when impact happens at low velocities.
Stress-strain curve can be divided into three regions [9] which is presented in Figure 10:
1. Linear elastic behaviour
2. Progressive crushing (cells buckle, yield or fracture) 3. Rapid stress increase (due to densification)
Figure 10: The three regions of the stress-strain curve.
3
2.3 Deformable barriers and their variants
Barriers made of aluminium honeycomb structure are often used to examine deformation and deformation patterns in static and dynamic impact tests. Designs of various shapes and crush strength properties have been developed to conform to the characteristics of the relevant test.
This paragraph presents different types of barriers and their usage in the industry.
In the development and implementation of these barriers there are some key organisations worth to mention, which will also be referred to in the following sections. European Enhanced Vehicle-safety Committee, EECV, is an organization consisting of representatives from several European nations. Working Groups, consisting of various specialists, conduct research within different fields related to vehicle safety. They advise the Committee who in turn discuss with appropriate authorities what actions to take in order to further ensure vehicle safety [10].
European New Car Assessment Program, Euro-NCAP, preform various tests of passenger cars in order to provide consumers with safety performance assessment. As a result, Euro-NCAP has become an important part in the drive for improved safety in passenger cars [11]. National Highway Traffic Safety Administration, NHTSA, is part of the United States Department of Transportation with the aim to reduce crashes and improve safety [12]. As part of their work to ensure safety, test protocols are developed and implemented including test equipment and details of test setup. The New Car Assessment Program, NCAP, was founded by NHTSA and is what the Euro-NCAP was modelled after [13]. The Insurance Institute for Highway Safety, IIHS, is an non-profit organisation initiating research within different areas of vehicle safety in order to reduce losses from vehicle crashes [14].
2.3.1 Types of barriers
Depending on the type of crash scenario tested, different barriers are used to represent the impacting vehicle. Typically, the barriers are divided into whether they are movable, for example placed on a trolley, or in a fixed position. The main difference being the barrier acting as the impacting object or the object being impacted respectively. Most of the movable barriers are used in side and rear impact testing while the fixed are used for frontal impact testing.
2.3.2 Mobile Progressive Deformable Barrier
The Frontal Impact and Compatibility Assessment Research Project, FIMCAR Project, initiated by the European Commission within the 7th Framework Program and founded in October 2009, is a project with the aim to asses and contribute to compatibility of vehicles in frontal collisions. The main objective is to propose assessment approach for frontal impacts. In the IX-MDB Test Procedure Protocol [15], the Moving Progressive Deformable Barrier, MPDB, was examined in a frontal crash in order to produce a protocol for moving deformable barriers in test procedures, assessing frontal compatibility. The MPDB, a further development of the Progressive Deformable Barrier, PDB, is constructed for use with a trolley (and thus, adding the mobile part of the barrier). It consists of three different blocks, of which the rear and front blocks deform consistently. The middle block has been designed to give a progressively increasing level of force with increasing deflection, giving the barrier its name. An aluminium back plate, bonded to the rear block, is used to mount the barrier on the trolley. Each block is bonded to an aluminium sheet, and the barrier is covered by a
4
cladding sheet, also aluminium, which is riveted on the face of the front block. The MPDB will be implemented in the Euro-NCAP protocol for frontal impact testing starting in 2020, Figure 11 presents the barrier mounted on the trolley with relevant dimensions. For image of the barrier see Figure 13 further down in this section.
Figure 11: View of MPDB and trolley to be implemented in the Euro-NCAP protocol [16].
In a study conducted at ISUZU Advanced Engineering Centre in Japan, the effectiveness of energy absorbing frontal underrun protection on trucks, in terms of fatality reduction, was investigated. The MPDB was used and further developed to suit a frontal impact crash with a truck [17]. In the study the use of a MPDB was motivated by cost reduction and generalization purposes. The properties of the crash test evaluated in the study includes an overlap of 50% and speed of MPDB at impact set to 80 km/h. Two major changes were implemented. Firstly, the height of the barrier was adjusted with the motivation that the upper part of the initial barrier might meet the stiff longitudinal members of the truck, tilt mechanism and cabin floor. If this were to happen in a full-scale vehicle-to-truck frontal crash, it would occur at a later stage. In a study from ADAC the same change was suggested [18], and the height was reduced by 133 mm (from the initial height of 700 mm). The second change regarded the issue of the front structure of the barrier. In a crash where the truck’s FUP acts as the main front structure, there is a risk of the MPDB bottoming out. The study proposed a solution of adding bumper structure from the Offset Deformable Barrier, ODB, on the MPDB with the function of spreading the local load from the single FUP into the barrier. The ODB is briefly explained in following paragraph. Figure 12 presents the initial PDB and the modified MPDB used for the truck-to-car crash test. In the figure a side view of the barrier is illustrated, the first image with the original height of 700 mm (with 150 mm ground clearance). In the second image the adjusted PDB has had its height lowered by 133 mm. In the final MPDB a bumper is added, here visualised as the red part.
5
Figure 12: Final MPDB used in the ISUZU study [17].
The ODB is a barrier developed between 1990 and 1995 by EECV WG 11 (working group 11) and introduced in the Euro NCAP in 1997, where it has been upgraded since 2015 [19].
It is used as part of a crash test which aims to evaluate the vehicle structure of the impacting car. The test represent a full-scale vehicle crash between two passenger cars of the same weight, traveling at a speed of 50 km/h. The setup consists of the barrier in a fixed position and the car (being tested) traveling with a speed of 64 km/h, crashing into the barrier with a 40% overlap. The barrier consists of two deformable parts; one main block with a crush strength of 0.342 MPa and a bumper structure with crush strength 1.72 MPa. The total dimension of the ODB is 1000 mm wide, 650 mm high and 540 mm deep. In order to enable mounting for the barriers fixed position, there is a backing sheet of 2 mm thickness, at which the barrier can be attached using bolts. The following images illustrates the barrier on its own, to the right in Figure 13, compared with that of MPDB, to the left in Figure 13. In Figure 14 examples of the barriers in use are presented.
Figure 13: The honeycomb structure of the ODB [20], left, and MPBD [21], right.
Figure 14: The ODB and MBD [22] in usage, left and right respectively.
6 Main properties of the MPDB:
Verified for impact velocity 64 km/h.
Total mass of trolley and barrier is 1400 kg.
Dimension of barrier, see Figure 15 to the right.
Crush strength of rear block: 1.72 MPa.
Crush strength of middle block: according to Euro NCAP technical specification.
Crush strength of front block: 0.342.
2.3.3 Advanced European Mobile Deformable Barrier
The Advanced European Movable Deformable Barrier, AE-MDB, is a further development of the MDB defined in the EEC Reg.95. It was initiated as a result of updated crash tests, with the aim for evaluating and improving compatibility between vehicles. The barrier is used mainly in side impact crash tests. The initial MDB consists of 6 blocks of aluminium honeycomb covered by a 0.5 mm aluminium plate. There are four different versions of the AE-MDB developed, all with small changes from the previous one. The initial AE-MDB was constructed with an increase in weight, from 950 to 1500 kg, as well as an increase in impact speed to 50 km/h. The edges on each side block were chamfered by 45 degrees, and the width of the side blocks increased by 100 mm, or to 600 mm. In order to better represent the car structure, the stiffness of the lower side blocks was increased. In 2002 the second version of the AE-MDB was developed. In order to follow the development in design of passenger cars, a bumper structure was added with a crush strength of 1.86 MPa, and with an aluminium plate of 3 mm thickness between the bumper and main blocks. The three lower blocks (D, E and F in Figure 16) were modified to increase the stiffness, with the purpose to ensure more uniform distribution. Development of the barrier continued as a part of the European 6th Framework project, APROSYS. In the third version, AE-MDB v3.1, an additional plate of aluminium was added to the bumper in order to prevent the thinner plate from breaking. The 3.1 version was shown to be a better representation of permanent deformation at 50 km/h than previous. The fourth and current version (used in Euro-NCAP from 2015), AE-MDB v3.9, had the stiffness of the outer blocks reduced by 55 % compared to v2. According to Cellbond the v3.1 and v3.9 gives more accurate representation of deformation and energy absorption in the door of the impacted vehicle (however, there are few differences in the injuries of the test dummies). Furthermore, the two versions are more representative of modern cars than previous versions and the ECE R95 Advanced 2000 barrier, in tests produced by APROSYS. Figure 16 presents the AE-MDB v3.9 barrier as well as the barrier in exploded isometric view, to the right.
Figure 15: Dimension of blocks of the MPDB, width of barrier: 1000 mm [31].
7
Figure 16: The fourth version of the AE-MDB, called AE-MDB v3.9 [23].
The European AE-MDB was evaluated and further developed in a study performed with the aim to design a family of barriers used as representative structures in frontal collisions against large-sized heavy vehicles. The usage of the current MDBs are not valid in high energy collisions as the structure will deform beyond the barrier, i.e. the depth of the deformation will exceed that of the barrier. This can be observed in the comparison of different data test from the NHTSA database: in crash test with passenger cars at 56 km/h, towards a rigid wall, the maximum deformation reached exceeds the length of the deformable barrier.
In the study, two new blocks of aluminium were added behind the deformable block of the AE-MDB, of crush strength 1.72 MPa and 0.62 MPa with respective length 150 mm and 400 mm. The elongation of the barrier decreases the risk of the barrier collapsing under more severe collisions. Results from the simulations show the deformation of the barrier to occur in a progressive way, meaning the rear block does not start to deform until the first block is almost completely deformed. This type of behaviour is consistent with observations of full-scale vehicles. The progressive deformation allows for the barrier to also be used in collisions of lower intensity.
Main properties of the AE-MDB:
Velocity at impact 50 km/h
Total mass of trolley and barrier is 1500 kg.
Dimension: in total 1700±2.5 mm x 570±1 mm x 500±2 mm.
Dimensions of each block: see Figure 17 to the right.
For crush strength of blocks A-E, see the AE- MDB Specification from Euro-NCAP [23].
Crush strength of bumper: 1.69 MPa.
Figure 17: Dimensions of each block of the AE-MDB [18].
8 2.3.4 IIHS Side Impact Barrier
Due to the large number of SUV and small off-road vehicles sold in the United States [24], the Insurance Institute for Highway Safety, IIHS, developed a barrier to represent these types of vehicles in crash tests. It is a continuation on the existing FMVSS (US Federal Standard for Side Impact Protection) 214 barrier, developed by NHTSA, and the first barrier to be made into an official standard. It is mainly used in tests evaluating passenger safety in rear and side impact collisions. The barrier consists of one main body, with crush strength of 0.31±0.017 MPa, and a bumper placed on the lower front of the main body, with crush strength 1.69±0.10 MPa. In total the dimensions are 1676 x 483 x 759 mm with tolerances
±6 mm for all sides [25]. The IIHS barrier shares several properties with its predecessor while certain important changes were made. The height of the IIHS barrier increased to 759 mm and the sides were chamfered with the aim to better represents vehicles such as SUVs and pickups. In comparison the impact face of the FMVSS 214 barrier is flat. The weight increased from 1361 kg to 1500 kg. It has been verified for an impact speed of 50 km/h.
Figure 18 presents an image with the FMVSS 214 barrier to the left, and the IIHS barrier to the right.
Figure 18: IIHS barrier to the left [26] and FMVSS 214 barrier to the right [27].
Main properties if the IIHS barrier:
Verified for impact speed of 50 km/h.
Total mass of trolley and barrier is 1500 kg.
Crush strength of main block is 0.31 MPa.
Crush strength of bumper (consisting of three honeycomb parts) is 1.69 MPa.
Main block is covered by 0.7 mm thick aluminium plates.
Bumper is covered, back and front, with 0.3 mm think aluminium plates.
Barrier has width of 1676 mm and height of 860 mm.
Trolley has track width of 1880 mm and wheel base of 2591 mm.
Compared to the trolley used in FMVSS 214, the front mounting plate is raised 100 m higher of the ground and has been extended 200 mm (resulting in the top surface of the mounting plate 300 m higher than the FMVSS 214 trolley). Figure 19 illustrates the complete setup: trolley and the IIHS barrier.
Figure 19:IIHS Barrier with trolley [34].
9 2.3.5 Oblique Mobile Deformable Barrier
The Oblique Mobile Deformable Barrier, O-MDB, is one of the largest and stiffest barrier currently available. As with the IIHS barrier, detailed in previous section, the O-MDB represents crash tests involving larger passenger cars such as SUVs and pick-ups. It was developed to be used mainly within the area of oblique offset impact testing. Figure 20 shows the barrier with trolley and Figure 21 shows a cut-through view with details of properties and its build-up.
Figure 20: Oblique Mobile Deformable Barrier, OMDB, with trolley (image provided by Plascore).
Main properties of the OBMD:
Verified for impact speed of 90 km/h.
Total mass of trolley and barrier is 2700 kg.
Crush strength of rear block is 1.71 MPa.
Crush strength of front block is 0.724 MPa.
The height of the barrier is 950 mm.
The width of the barrier is 2200 mm.
The depth of the barrier is 606 mm.
Figure 21: Properties of the OMDB (image provided by Cellbond).
10
3 IMPLEMENTATION
In this chapter the development of the test method is described. It includes the requirements, the selection of test setup and the representative structure for the passenger car and truck.
3.1 Requirements
In order to ensure the development of the simplified test rig will account for the important properties of the full-scale vehicle test setup, a requirement specification has been constructed.
It includes all the needs and wants. For example, impact velocity for this thesis has been chosen as 70 km/h, since the barriers are known to work up to 65 km/h. Therefore, further research is needed. See Table 1 below for further information.
Table 1. Requirement specification.
Requirement Specification Type
Impact velocity 70 km/h Need
Overlap of passenger car 75 % Need
Height from FUP to floor 475 mm (from centreline of FUP) Need Maximum height from FUP to floor 490 mm (from centreline of FUP) Want Lowest height from FUP to floor 300 mm (from centreline of FUP) Want
Range of adjustment from maximum height 100 mm Need
Maximum deformation of any FUP subsystem* 400 mm Need
Load of truck structure 10 ton Need
Mobility of truck structure Idle state Need
Dummy (-ies) Not required -
*Any point of the FUP subsystem must not be deformed more than specified behind the truck’s main point, after deformation. The requirement refers to the direction, x-direction of which the passenger is travelling along.
3.2 Concept generation and evaluation
In section 3.2.1 the usage of barrier and trolley was selected for further investigation. In order to continue the design of the new test method, different concepts in terms of setup of the car – and truck structures are constructed and evaluated. In the following sections two different concepts for setup are described and evaluated.
3.2.1 Concepts
In order to increase repeatability of the test and to reduce costs of a full-scale test simplifications have to be made. Focus is gathered on the front of the truck and the most important components have to be identified, which would act in a similar manner as the full-scale crash. Thus, two concepts were chosen for analysis, which would represent the crash test effects:
1) Trolley crashes into a fixed underrun protection, as seen in the Figure 22.
2) Moving trolley crashes into a stationary trolley (with mounted underrun protection) without velocity, as seen in the Figure 23.
11
Figure 22: First concept.
Figure 23: Second concept.
It is important to mention that the mass of the impacting body cannot be varied and must represent a typical European passenger car. For example Volkswagen Golf MK7 which weighs around 1500 kg. It is a must, because accelerometers will be mounted on a trolley which will provide g-values that represent injury levels on passengers.
3.2.2 Evaluation
Calculations were setup according to the simplified models where spring coefficient has been calculated assuming the spring is represented by 2.2 m long UPN-80 beam, see Figure 24. The dimensions are stated in Table 2.
Table 2. Beam measurements [12].
Beam type
Mass per metre
kg/m
Depth of section H, mm
Width of section
B, mm
Web Thickness
TW, mm
Flange thickness
TF, mm
Root radius
R1, mm
Toe radius
R2, mm
Depth between
fillets d, mm
Area cm2
UPN- 80
8.64 80 45 6.0 8.0 8.0 4.0 47 11.00
12
Equations (1) to (4) concerns the calculations used for evaluation of the first concept:
Model 1.
Kinetic – Strain energy principle
E
k U
(1)where 𝐸𝑘 – kinetic energy of the moving mass, 𝑈 – strain energy.
2 2
2 2
mv kx (2)
where 𝑚1 is the mass of an impacting body, 𝑣11 – velocity of the impacting body, 𝑘 – spring coefficient, 𝑥 – spring longitudinal deformation.
Displacement
mv2
x k (3)
Force generated by displacement
F = kx (4)
Equations (5) to (14) concern the calculations used for evaluation of the second concept:
Model 2.
Kinetic – Strain energy principle
1 2
k k
E E U
(5)where 𝐸𝑘1 – kinetic energy of the impacting mass, 𝐸𝑘2 – transmitted kinetic energy into the stationary 2nd body after collision, 𝑈 – strain energy.
2 2 2
1 0 2 2
2 2 2
m v m v kx (6)
where 𝑚1 – mass of the impacting body, 𝑣0 – initial velocity before the collision, 𝑣2 – velocity of the idle mass after crash, 𝑘 – spring coefficient, 𝑥 – spring longitudinal deformation.
Figure 24: UPN-80 beam cross-section [19].
13
Theory behind impact describes elastic or plastic collision through restitution coefficient - 𝑒. 0< 𝑒<1 where 0 stands for perfectly plastic and 1 – perfectly elastic [28].
1 2
1 0
1 2
(m e m )
v v
m m
(7)
where 𝑣1 – velocity of the impacting body after collision, 𝑚2 – mass of the idle body, 𝑒 – restitution coefficient.
1
2 0
1 2
(1 e m)
v v
m m
(8)
Kinetic energy after impact
2 1
1 2 1
0 0
k Tot
k k
E v
E if
E E v
(9)
Kinetic energy of the impacting body after collision
2 1 1
1 2
k
E m v
(10)
Kinetic energy of the idle body after collision
2 2 2
2 2
k
E m v (11)
Initial kinetic energy
2 1 0 k 2 E m v
(12)
Displacement
2(Ek ETot)
x k
(13)
Force
𝐹 = 𝑘𝑥 (14)
Within Table 3 values are described to define the problem. Inside Appendix C the Matlab code for the estimation can be found.
Table 3. Input values.
Parameter Value Unit Meaning
𝑚1 1500 kg Impacting body mass
𝑚2 150:500:10500 kg Stationary body mass
𝑣11 (30:2:70)/3.6 m/s Impacting velocity
𝑣12 0 m/s Second body velocity
𝑒 0:0.25:1 - Restitution
coefficient
14
Energy distribution graph, Figure 25, after impact shows that restitution coefficient (e) has influence on the overall system response. When e = 1 the structure behaves perfectly elastic and when e = 0 system acts in perfectly plastic case. However, a more realistic case would be e = 0.25. Figure 25 and Figure 26 shows Energy ratio (𝐸𝑇𝑜𝑡/𝐸𝑘) versus body mass ratio or how energy tends to distribute in regard to time.
Figure 25: Energy distribution after impact.
Figure 26: Energy ratio vs time.
From Figure 27 and Figure 28 it can be observed that force tends to rise after initial contact and that the difference regarding elasticity is very small. Additionally, the biggest displacement values are achieved when the structure is in perfectly plastic state,
15
because then the highest amount of absorbed energy is transformed into deformation.
Only, SDOF (first concept) experiences higher displacement, because it experiences higher force values. The biggest force value for second concept is around 6.9 MN, when the system is in perfect elastic mode and of 5.5 MN are generated during perfectly plastic state.
Figure 27: Force vs displacement.
Figure 28: Force vs time.
16
Restitution coefficients influence on the system is represented within Table 4. It shows how plasticity of a structure has a direct correlation to displacement occurring during deformation of a structure.
Table 4. Maximum displacement at roughest conditions (>60 km/h, >10000 kg)
Restitution
coefficient, e 0 0.25 0.5 0.75 1
Maximum displacement,
mm
71.6 69.7 66.3 62.1 56.9
Both concepts have been compared as shown in Figure 29. The first concept (SDOF) is a perfect elastic case and the initial kinetic energy is transformed into displacement or heat. However, latter is not considered. The highest displacement occurs at 76.7 mm and the force at that point is 7.4 MN. The value is higher when compared to a two body system and it is important to stress that the first concept doesn’t take into account plasticity. Therefore, displacement would increase if plastic behaviour of a body would be introduced.
Figure 29: Force vs displacement. Comparison between two models.
In order to find out what influence velocity has on the two body system every other variable was set as a specific value except restitution factor. Velocity has been varied from 30-70 km/h. The difference is almost negligible as shown in Figure 30 and Figure 31.
17
Figure 30: Force vs displacement (velocity impact).
Figure 31: Force vs time (velocity impact).
In order to find out what influence truck mass has on the two body system every other variable was set as a specific value except restitution factor. Mass has been varied from 150-10500 kg with a step of 500 kg. Truck mass has an impact on the force acting in the system when compared to the influence from the velocity, as can be seen in Figure 32 and Figure 33.
18
Figure 32: Force vs displacement (mass variation).
Figure 33: Force vs time (mass variation).
19 3.2.3 Selection
Two models have been evaluated. Since the mass of the impacting body cannot be varied and must represent the value of a typical European family car (Volkswagen Golf) there is not a lot of space to play around with it. Considering that the velocity will affect strain rate directly due to the behaviour of the structure differentiating when comparing low-velocity to mid/high-velocity cases. Therefore, there is also little room to play around with the velocity as well.
The first concept experiences higher deformation than the second one, which is represented by displacement. As mentioned previously, first concept does not take into consideration elasticity and the system is regarded as perfectly elastic. Therefore, if plastic properties would be introduced the difference in deformation between models would increase. Nonetheless, the full scale test does not use fully fixed structures in a crash. Thus, the second concept has been chosen to be evaluated further on.
Additionally, it is important to stress, that these calculations have been made for comparison between two cases of which one represents a more realistic behaviour.
However, they are oversimplified just for the concept generation phase. Results cannot be taken as an indication of mechanics behind a crash. The structures do not behave in a linear manner during plastic deformation. Besides, such factors as buckling, strain rate, shear stress/rate, etc. are not taken into consideration. Therefore, there is a need to check the model numerically using finite element analysis.
3.2.4 Test-rig setup
Few options were considered during concept generation phase. One of the first discussed ideas was to use a hydraulic actuator and with a specified force apply a load on the front underrun protection. However, as the data within calculations suggest this type of test methodology would not create deformations that occur during a collision and would merely suffice for material strength estimations. Thus, another concept was considered, where an impact barrier would be mounted on a platform that moves on rails and crashes into a stationary or idle object, which has underrun protection mounted on it. On the other hand, this setup requires tracks and a custom facility where they can be used. Moreover, the mobility is restricted by railway tracks and objects will not be able to represent a real case scenario during an impact. Therefore, a trolley-to-trolley concept has been chosen after careful investigation.
To represent the passenger car a barrier is mounted on a trolley. This type of testing has become a standard within the car industry and is well implemented in crash testing.
Several different types of barriers have been developed over the years, enabling tests of frontal, oblique, side and rear impacts. Furthermore, the barriers are continuously developed and improved to meet the changes in car designs and to ensure accurate representation of a modern car’s energy absorption during impact. The research and development indicates the barriers will continue to be the primary testing structure, even further implemented and thus increasing the potential of the test rig.
The selected test rig setup will be as follows: One trolley is equipped with a barrier (representing the passenger car) while the other trolley holds the FUP as well as additional components. The trolley with the barrier is launched into the other one, which stands idle, at a certain velocity.
20
3.3 Selection of barrier
In this section the evaluation and selection of barrier is presented. As mentioned previously, it has been decided the test setup will represent the passenger car via a barrier mounted on a trolley. The trolley will be launched and crash into the structure representing the truck at a certain velocity. In order to decide appropriate barrier, inputs from suppliers and test facilities have been regarded. The suppliers contacted and consulted include Cellbond, Argosy International and Plascore. Test company TASS International have also been consulted.
3.3.1 AE-MDB
Cellbond recommended the AE-MDB. The barrier is designed to be representative of a standard European passenger car, the weight is close to that of the car used in the full- scale test and the barrier is well implemented in current testing (part of Euro-NHTSA test protocol). However due to its usage in side impact testing the front structure of the barrier is made with chamfered sides, which is of concern (as it is not a flat structure which provides a more accurate representation in frontal impact testing). Furthermore, it is tested and verified for an impact velocity of 50 km/h, in comparison to the required velocity of 70 km/h.
3.3.2 MPDB
The MPDB was recommended by Argosy International as well as mentioned by representative from TASS. As with the AE-MDB the barrier is made to fit European standards in crash testing, it has a weight of 1400 kg which is very close to that of the Golf used in the full-scale test. The barrier is one of the latest developments of barriers and will be implemented in the Euro-NCAP frontal impact testing beginning in 2020. It is verified for impact speed at 64 km/h, making it one of the evaluated barriers with highest velocity. The concern regards the dimensions of the barrier, more specifically the width. As it has a width of 1000 mm, the requirement of 75% overlap cannot be fulfilled. In order to meet the demand of the overlap, the barrier must be approximately at least 1350 mm in width.
3.3.3 IIHS Barrier & O-MDB
Argosy International as well as Cellbond also recommended the IIHS Barrier. The barrier is similar in weight and velocity as with the two previously mentioned ones, 1500 kg and 50km/h, and the honeycomb blocks share similar crush strength properties.
However, the size of the barrier is distinctively larger, which is due to the barrier being specifically produced to represent larger vehicles such as pick-ups and SUVs. This becomes an issue as the test rig aims to replicate a test with a smaller vehicle. In addition, as with the AE-MDB, the barrier is used for side impact testing and does not have a flat front structure.
Another barrier suggested by Cellbond and Plascore, is the O-MDB. It is one of the most large and stiff barriers, used in tests with impact velocities at 90 km/h. The barrier weighs 2700 kg and the dimensions are, as with the IIHS barrier, made to represent vehicles much larger than the one used in the full-scale vehicle crash tests. It has a flat front structure and is used in frontal impact tests. The weight and the size eliminate this barrier as a possible alternative.
