- Email: [email protected]
Computational and experimental crash analysis of the road safety barrier
Engineering Failure Analysis 12 (2005) 963–973 www.elsevier.com/locate/engfailanal
Computational and experimental crash analysis of the road safety barrier Z. Ren *, M. Vesenjak University of Maribor, Faculty of Mechanical Engineering, Smetanova ul 17, SI-2000 Maribor, Slovenia Received 12 July 2004; accepted 14 December 2004 Available online 7 April 2005
Abstract The paper describes the computational analysis and experimental crash tests of a new road safety barrier. The purpose of this research was to develop and evaluate a full-scale computational model of the road safety barrier for use in crash simulations and to further compare the computational results with real crash test data. The impact severity and stiffness of the new design have been evaluated with the dynamic nonlinear elasto-plastic analysis of the three-dimensional road safety barrier within the framework of the finite element method with LS-DYNA code. Comparison of computational and experimental results proved the correctness of the computational model. The tests have also shown that the new safety barrier assures controllable crash energy absorption which in turn increases the safety of vehicle occupants. 2005 Elsevier Ltd. All rights reserved. Keywords: Safety barrier; Roadside safety; Computational crash analysis; Experimental crash test
1. Introduction One of the major problems in road transportation is to assure adequate safety level for road users. To maintain and improve road safety, it is often necessary to install certain devices that are intended to restrain vehicles and pedestrians from entering dangerous areas. The road safety barriers that are designed according to the European EN 1317 standard provide certain levels of vehicle containment, properly redirect errant vehicles back on the road and provide guidance for pedestrians and other road users. To provide appropriate safety levels for impacting vehicle occupants, the safety barriers (Fig. 1) should be designed so as to absorb as much impact energy as possible through their deformation and at the same
*
Corresponding author. Tel.: +386 2 220 7702; fax: +386 2 220 7990. E-mail addresses: [email protected] (Z. Ren), [email protected] (M. Vesenjak).
1350-6307/$ - see front matter 2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.engfailanal.2004.12.033
964
Z. Ren, M. Vesenjak / Engineering Failure Analysis 12 (2005) 963–973
Fig. 1. Currently used road safety barrier in Slovenia.
time maintain their integrity. Practical observations of installed systems along SlovenianÕs highways, which have not changed much in the past several decades, indicate that the current design of the road barrier is too stiff, which results in unacceptable decelerations during vehicle impact. By adopting the European transportation legislation, it was necessary to re-evaluate the safety barriers and propose certain design changes to increase the safety level. Behaviour of the new safety barrier designs under test vehicle impact conditions was simulated by using the dynamic nonlinear finite element code LS-DYNA. ¨ V Automotive test facility in Germany Additionally, a full scale crash tests were performed at the TU where the new road safety system design was tested according to EN 1317 standard.
2. Roadside safety Roadside safety addresses an area outside of the roadway and is an important component of the total roadway design. From a safety perspective, the ideal highway has roadsides and median areas that are flat and unobstructed by hazards. Elements such as side slopes, fixed objects and water are potential hazards which a vehicle might encounter when it leaves the roadway. These hazards present varying degree of danger to vehicle and its occupants. Unfortunately, geography and economics do not always allow ideal highway conditions. The mitigative measures to be taken depend on the probability of an accident occurring, the likely severity, and the available resources [1]. Road design in the future will increasingly focus on the safety of all users. Upgrading existing roads to higher safety standards in a new road construction, coupled with greater driverÕs awareness of factors which can cause accidents, will lead to significant savings in road trauma and crash costs. Road maintenance such as resurfacing to improve skid resistance will continue to be a priority [2]. To achieve safer roads one has to consider three key areas: Safer people: to encourage safe behaviour by ensuring that the drivers adhere to the speed limits, that they do not drive if impaired by alcohol or fatigue, that all vehicle occupants are using seatbelts and that young drivers have adequate knowledge and supervision.
Z. Ren, M. Vesenjak / Engineering Failure Analysis 12 (2005) 963–973
965
Safer vehicles: continuing improvements in vehicle safety with introduction of advanced active and passive safety devices continues to be a priority in automotive industry. Current developments in vehicle engineering ensure that all systems within a vehicle are combined to provide optimum safety levels. Safer roads: to lower the speed limits, to build and maintain better and safer roads, to improve safety barriers, to build traffic priority systems to ensure quicker responses by emergency services. Current improvements in road building legislation ensure that all new roads are built to the most stringent safety standards.
3. Safety barriers and the EN 1317 standard The road safety barriers installed on highways in Europe need to fulfil the EN 1317 standard in terms of the impacting vehicle containment level, the level of expected vehicle occupant decelerations during impact and the consequent barrier system deformation [3]. The standard prescribes criteria which the safety barrier has to fulfill under specific impact conditions. There are many different parameters that have to be taken into account when a vehicle collides with the safety barrier: vehicle velocity (v), vehicle mass (m), impact angle (a), type and behaviour of a vehicle and road conditions (Fig. 2). The safety barriers have to sustain impact of different vehicle types (from passenger cars to trucks) under different impact conditions regarding the vehicle velocity, impact angle and road conditions. In case of a lower-weight vehicle (passenger car) impact, the restraint system should possess the ability to deform, so that the kinetic energy of an impact is absorbed mostly by the barrier and vehicle deformation. This significantly reduces deceleration levels experienced by vehicle occupants and increases their safety. However, in a case of higher-weight vehicle (truck, bus) impact, the system should contain and redirect the vehicle back on the road without complete breakage of the principal longitudinal elements of the system. Thus, the safety barrier design is a compromise between its stiffness (deformability) and strength. The road safety barriers have to fulfill the following criteria according to EN 1317-2: Containment level – represents the level of containment for different types of vehicles. The standard defines four major classes of containment levels (low angle, normal, higher and very high containment), which are further subdivided (T1, . . . ,H4b). Impact severity – a measure of impact consequences for the vehicle occupants. Three measures are used: the acceleration severity index (ASI), the theoretical head impact velocity (THIV) and the post-impact head deceleration (PHD). Deformation of the barrier – a working width of the barrier, which is a distance between the side of the W-beam facing the traffic before the impact and the maximum lateral position of any major part of the system after the impact. There are eight classes of deformation of the restraint system (W1 – 0.6 m, . . . ,W8 – 3.5 m).
m v
Fig. 2. Parameters of a theoretical vehicle impact.
966
Z. Ren, M. Vesenjak / Engineering Failure Analysis 12 (2005) 963–973
Average vehicle impact severity is usually measured with evaluation of the maximum ASI, which is defined as ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi s 2 2 2 ax ay az ASIðtÞ ¼ þ þ < 1; ð1Þ b b b ax ay az y ¼ 9 g; where ax ¼ 12 g; a az ¼ 10 g are limit values of the coordinate acceleration and ^ax ; ^ay and ^az are the components of the averaged acceleration of the vehicle mass centre over a moving time interval of d = 50 ms. If maximum ASI value exceeds 1.0 or 1.4 in some cases, then it is considered that the consequences of an impact are dangerous or even lethal for the occupants of impacting vehicle.
4. Road safety barrier The barriers are usually made of steel, although higher containment levels can sometimes be achieved only by using concrete blocks. The evaluated road safety barrier (Fig. 3) is made of construction steel S 235 (St 37-2 according to DIN) and comprises a W-shaped guardrail, distance spacer and post rammed into the soil to the depth corresponding to 2/3 of its length. The W-shaped guardrail is made of 3 mm thick metal sheet with improved strength characteristics. The usual length of the W-guardrail segments is 4200 mm and the splice length equals 200 mm. The guardrail segments are longitudinally connected by bolts. The distance spacer is hexagonal shaped, made of 4 mm thick metal sheet and measures 260 mm in length and 220 mm in height [4]. The post is C-shaped with dimensions 55 mm · 100 mm · 4 mm and is 1900 mm long. The distance between posts depends on the required containment level and can be equal to 1.33, 2 or 4 m. Posts are always oriented with the closed profile face towards the traffic flow direction. W-guardrail and distance spacers are joined with M16 bolts of strength class 5.8, while the distance spacer and the posts are joined with M10 bolts of the same strength class. The later are supposed to fail in the early stages of the crash so that the guardrail is released from the post instead of being pulled down with it.
Fig. 3. Elements of the road barrier.
Z. Ren, M. Vesenjak / Engineering Failure Analysis 12 (2005) 963–973
967
5. Description of the computational model A 34 m long segment of the road safety barrier was found to be sufficient for conducted impact analyses. The guardrail was modelled as a continuous element under presumption that the guardrail splices were not the weakest link, according to Engstrand [5]. The model comprised also 9 posts separated by the distance of 4 m. All component parts were modelled with the 4-node Belytschko-Tsay shell elements with three integration points through the shell thickness [6,7]. Thicknesses of model parts are the following: guardrail 3 mm and distance spacer and post 4 mm. The impacting vehicle, type Metro (publicly accessible on web page: http://www.ncac.gwu.edu), was modelled mostly by shell elements and its weight was tuned to prescribed 900 kg (Fig. 4) [8]. 5.1. Material model All parts of the road restraint system are made of the construction steel S 235. Standard tensile tests of different thickness specimens were carried out according to the standard DIN 50 115 to obtain realistic material properties of analyzed road restraint elements in terms of their YoungÕs modulus, yield stress, elasto-plastic relationship and ultimate strength. Obtained material parameters were used to define isotropic bilinear elasto-plastic material model with kinematic hardening in subsequent computational analyses. Previously experimentally determined dynamic material response parameters were considered in all computational analyses. It was presumed that the bolts and welds are rigid. The level of effective plastic strain was used as a failure condition in the model. The maximum effective strain was set to 0.28 which corresponds to 28% ductility of S 235 steel. When the effective plastic strain reached that value, the element load carrying capability (stress) was reduced to zero, effectively removing the element from the model. 5.2. Boundary conditions and contact description The posts were fixed with viscoplastic springs (Fig. 5) in order to simulate the soil influence. Linear springs were attached at both ends of the guardrail (Fig. 5) to simulate its continuation. All surfaces of the model were defined as one contact group, thus, effectively accounting for multiple selfcontacting regimes during computational impact analyses. The static and dynamic friction coefficient between all parts was set to 0.1 and 0.05, respectively.
Fig. 4. Impacting vehicle.
968
Z. Ren, M. Vesenjak / Engineering Failure Analysis 12 (2005) 963–973
Fig. 5. Boundary conditions.
5.3. Dynamic analysis properties The vehicle was prescribed to have the initial velocity of 100 km/h at an impact angle of 20 with regard to the guardrail (Fig. 5). The analysis time interval was set to 350 ms, with results output required every 1 ms. The time step for explicit transient dynamic analysis was set to 1.4 ls by the used code LS-Dyna PC Version 970 with regard to the lowest resonant frequency of the structure. The analyses run 85– 95 h on a PowerPC with 2 AMD 2000+ MHz processors and 2 GB RAM, depending on the complexity of the individual model.
6. Analysis of computational results The results of computational simulations are shown in Figs. 6–9. Fig. 6 shows the deformation steps of the safety barrier and appropriate containment of the impacting vehicle. The computed contact length between the barrier and the vehicle was 10.8 m. Computational simulations have shown that extreme decelerations of the vehicle mass centre occur during initial vehicle impact at the guardrail and later when the vehicle hits the post. ASI index was well bellow the safety limit and was the highest when the front wheel hit the first post (max ASI = 0.66), see Fig. 13. Simulations have also clearly proved that appropriate distance spacer design, through its controlled deformation, significantly reduces the decelerations experienced by vehicle occupants in the initial phase of an impact. The hexagonal distance spacer deforms in a controllable manner until the guardrail is pushed to the post (Fig. 7) and absorbs the highest possible amount of crash energy. From Fig. 8, it is also obvious that a deceleration peak occurs when the front wheel hits the post. With a guidance profile those high decelerations can be avoided. The simulations have also shown that the weak connection between the distance spacer and the post failed as designed. The connection failed just before the vehicle hit the first post (Fig. 9). The guardrail was released and consequently contained the vehicle on the road. The working width (maximum lateral dynamic deformation) of the simulated safety barrier was equal to 1.2 m, which corresponds to working width class W4 according to EN-1317-2.
Z. Ren, M. Vesenjak / Engineering Failure Analysis 12 (2005) 963–973
969
Fig. 6. Computational simulation of the vehicle impact.
7. Full-scale crash test ¨ V Automotive test facility in GerThe new road safety barrier was tested in a full scale crash test at TU many according to the EN 1317 standard. The crash test was carried with the FIAT Uno as the impacting vehicle with the same initial impact conditions as those used in the simulations (Fig. 10). The road restraint system performed well and was successful in redirecting the vehicles back on the road without collapsing (Figs. 11 and 12). The measured maximum ASI index was equal to 0.63 (Fig. 13), the working width was 0.95 m (class W3 according to EN-1317-2) and the length of the contact between the impacting vehicle and the barrier was equal to 12 m.
970
Z. Ren, M. Vesenjak / Engineering Failure Analysis 12 (2005) 963–973
Fig. 7. Distance spacer deformation.
Fig. 8. Wheel impact at the post.
Fig. 9. Weak connection between the distance spacer and the post.
Z. Ren, M. Vesenjak / Engineering Failure Analysis 12 (2005) 963–973
Fig. 10. Crash test vehicle (FIAT Uno) (a) before and (b) after the impact.
Fig. 11. Test vehicle impact at the road restraint system.
Fig. 12. Deformation of the road restraint system.
971
972
Z. Ren, M. Vesenjak / Engineering Failure Analysis 12 (2005) 963–973 1.00 0.90 computational results 0.80
experimental results
0.70
ASI
0.60 0.50 0.40 0.30 0.20 0.10 0.00 0.00
0.05
0.10
0.15
0.20
0.25
0.30
0.35
Time [s]
Fig. 13. Computational and experimental results comparison.
8. Comparison of the computational and experimental results From comparison of computational (Section 6) and experimental (Section 7) results it can be observed that the results compare very well in relation to the working width of the system (computed 1.2 m; measured 0.95 m) and contact length (computed 10.8 m; measured 12 m). The differences can be attributed to parameters used to describe material dynamic behaviour in computations, which were obviously underestimating the stiffness increase of the material under dynamic impact loading. Computational and experimental ASI index histories are shown in Fig. 13. ASI index for the tested system does not exceed the limiting value in any case. The highest values (computational 0.66; measured 0.63), are observed when the front wheel hits the first post. 9. Conclusions Computational nonlinear explicit dynamic analysis was employed for evaluation of the new road safety barrier behaviour under test vehicle impact conditions. Computational simulation predictions have been compared with the results of the full-scale crash test. The used computational model provided comparable results to experimental measurements and can thus be used for computational evaluation of other road safety barriers in order to avoid numerous expensive full-scale crash tests. The tests have also shown that the new safety barrier assures controllable crash energy absorption which in turn increases the safety of vehicle occupants. References [1] Washington State Department of Transportation. Roadside safety-design manual, Washington; 2003. [2] Road Safety 2010 – A Framework for Saving 2000 Lives by the year 2010. Roads and Traffic Authory, New South Wales.
Z. Ren, M. Vesenjak / Engineering Failure Analysis 12 (2005) 963–973 [3] [4] [5] [6] [7] [8]
973
European Committee for Standardization. European Standard EN 1317-1, EN 1317-2, Road Restraint Systems; 1998. Vesenjak M, Ren Z. Computer-aided design of a road restraint barrier. J Mech Eng 2003;49(10):509–19. Engstrand KE. Improvements to the weak – post W – Beam Guardrail. Worcester: Worcester Polytechnic Institute; 2000. Livermore Software Technology Corporation, LS-DYNA keyword userÕs manual; 2001. Livermore Software Technology Corporation, LS-DYNA theoretical manual; 1998. Vesenjak M, Ren Z. Dynamic analysis of a road restraint systemÕs deformation resulting from a vehicle impact. J Mech Eng 2003;49(12):586–92.
Computational and experimental crash analysis of the road safety barrier Z. Ren *, M. Vesenjak University of Maribor, Faculty of Mechanical Engineering, Smetanova ul 17, SI-2000 Maribor, Slovenia Received 12 July 2004; accepted 14 December 2004 Available online 7 April 2005
Abstract The paper describes the computational analysis and experimental crash tests of a new road safety barrier. The purpose of this research was to develop and evaluate a full-scale computational model of the road safety barrier for use in crash simulations and to further compare the computational results with real crash test data. The impact severity and stiffness of the new design have been evaluated with the dynamic nonlinear elasto-plastic analysis of the three-dimensional road safety barrier within the framework of the finite element method with LS-DYNA code. Comparison of computational and experimental results proved the correctness of the computational model. The tests have also shown that the new safety barrier assures controllable crash energy absorption which in turn increases the safety of vehicle occupants. 2005 Elsevier Ltd. All rights reserved. Keywords: Safety barrier; Roadside safety; Computational crash analysis; Experimental crash test
1. Introduction One of the major problems in road transportation is to assure adequate safety level for road users. To maintain and improve road safety, it is often necessary to install certain devices that are intended to restrain vehicles and pedestrians from entering dangerous areas. The road safety barriers that are designed according to the European EN 1317 standard provide certain levels of vehicle containment, properly redirect errant vehicles back on the road and provide guidance for pedestrians and other road users. To provide appropriate safety levels for impacting vehicle occupants, the safety barriers (Fig. 1) should be designed so as to absorb as much impact energy as possible through their deformation and at the same
*
Corresponding author. Tel.: +386 2 220 7702; fax: +386 2 220 7990. E-mail addresses: [email protected] (Z. Ren), [email protected] (M. Vesenjak).
1350-6307/$ - see front matter 2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.engfailanal.2004.12.033
964
Z. Ren, M. Vesenjak / Engineering Failure Analysis 12 (2005) 963–973
Fig. 1. Currently used road safety barrier in Slovenia.
time maintain their integrity. Practical observations of installed systems along SlovenianÕs highways, which have not changed much in the past several decades, indicate that the current design of the road barrier is too stiff, which results in unacceptable decelerations during vehicle impact. By adopting the European transportation legislation, it was necessary to re-evaluate the safety barriers and propose certain design changes to increase the safety level. Behaviour of the new safety barrier designs under test vehicle impact conditions was simulated by using the dynamic nonlinear finite element code LS-DYNA. ¨ V Automotive test facility in Germany Additionally, a full scale crash tests were performed at the TU where the new road safety system design was tested according to EN 1317 standard.
2. Roadside safety Roadside safety addresses an area outside of the roadway and is an important component of the total roadway design. From a safety perspective, the ideal highway has roadsides and median areas that are flat and unobstructed by hazards. Elements such as side slopes, fixed objects and water are potential hazards which a vehicle might encounter when it leaves the roadway. These hazards present varying degree of danger to vehicle and its occupants. Unfortunately, geography and economics do not always allow ideal highway conditions. The mitigative measures to be taken depend on the probability of an accident occurring, the likely severity, and the available resources [1]. Road design in the future will increasingly focus on the safety of all users. Upgrading existing roads to higher safety standards in a new road construction, coupled with greater driverÕs awareness of factors which can cause accidents, will lead to significant savings in road trauma and crash costs. Road maintenance such as resurfacing to improve skid resistance will continue to be a priority [2]. To achieve safer roads one has to consider three key areas: Safer people: to encourage safe behaviour by ensuring that the drivers adhere to the speed limits, that they do not drive if impaired by alcohol or fatigue, that all vehicle occupants are using seatbelts and that young drivers have adequate knowledge and supervision.
Z. Ren, M. Vesenjak / Engineering Failure Analysis 12 (2005) 963–973
965
Safer vehicles: continuing improvements in vehicle safety with introduction of advanced active and passive safety devices continues to be a priority in automotive industry. Current developments in vehicle engineering ensure that all systems within a vehicle are combined to provide optimum safety levels. Safer roads: to lower the speed limits, to build and maintain better and safer roads, to improve safety barriers, to build traffic priority systems to ensure quicker responses by emergency services. Current improvements in road building legislation ensure that all new roads are built to the most stringent safety standards.
3. Safety barriers and the EN 1317 standard The road safety barriers installed on highways in Europe need to fulfil the EN 1317 standard in terms of the impacting vehicle containment level, the level of expected vehicle occupant decelerations during impact and the consequent barrier system deformation [3]. The standard prescribes criteria which the safety barrier has to fulfill under specific impact conditions. There are many different parameters that have to be taken into account when a vehicle collides with the safety barrier: vehicle velocity (v), vehicle mass (m), impact angle (a), type and behaviour of a vehicle and road conditions (Fig. 2). The safety barriers have to sustain impact of different vehicle types (from passenger cars to trucks) under different impact conditions regarding the vehicle velocity, impact angle and road conditions. In case of a lower-weight vehicle (passenger car) impact, the restraint system should possess the ability to deform, so that the kinetic energy of an impact is absorbed mostly by the barrier and vehicle deformation. This significantly reduces deceleration levels experienced by vehicle occupants and increases their safety. However, in a case of higher-weight vehicle (truck, bus) impact, the system should contain and redirect the vehicle back on the road without complete breakage of the principal longitudinal elements of the system. Thus, the safety barrier design is a compromise between its stiffness (deformability) and strength. The road safety barriers have to fulfill the following criteria according to EN 1317-2: Containment level – represents the level of containment for different types of vehicles. The standard defines four major classes of containment levels (low angle, normal, higher and very high containment), which are further subdivided (T1, . . . ,H4b). Impact severity – a measure of impact consequences for the vehicle occupants. Three measures are used: the acceleration severity index (ASI), the theoretical head impact velocity (THIV) and the post-impact head deceleration (PHD). Deformation of the barrier – a working width of the barrier, which is a distance between the side of the W-beam facing the traffic before the impact and the maximum lateral position of any major part of the system after the impact. There are eight classes of deformation of the restraint system (W1 – 0.6 m, . . . ,W8 – 3.5 m).
m v
Fig. 2. Parameters of a theoretical vehicle impact.
966
Z. Ren, M. Vesenjak / Engineering Failure Analysis 12 (2005) 963–973
Average vehicle impact severity is usually measured with evaluation of the maximum ASI, which is defined as ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi s 2 2 2 ax ay az ASIðtÞ ¼ þ þ < 1; ð1Þ b b b ax ay az y ¼ 9 g; where ax ¼ 12 g; a az ¼ 10 g are limit values of the coordinate acceleration and ^ax ; ^ay and ^az are the components of the averaged acceleration of the vehicle mass centre over a moving time interval of d = 50 ms. If maximum ASI value exceeds 1.0 or 1.4 in some cases, then it is considered that the consequences of an impact are dangerous or even lethal for the occupants of impacting vehicle.
4. Road safety barrier The barriers are usually made of steel, although higher containment levels can sometimes be achieved only by using concrete blocks. The evaluated road safety barrier (Fig. 3) is made of construction steel S 235 (St 37-2 according to DIN) and comprises a W-shaped guardrail, distance spacer and post rammed into the soil to the depth corresponding to 2/3 of its length. The W-shaped guardrail is made of 3 mm thick metal sheet with improved strength characteristics. The usual length of the W-guardrail segments is 4200 mm and the splice length equals 200 mm. The guardrail segments are longitudinally connected by bolts. The distance spacer is hexagonal shaped, made of 4 mm thick metal sheet and measures 260 mm in length and 220 mm in height [4]. The post is C-shaped with dimensions 55 mm · 100 mm · 4 mm and is 1900 mm long. The distance between posts depends on the required containment level and can be equal to 1.33, 2 or 4 m. Posts are always oriented with the closed profile face towards the traffic flow direction. W-guardrail and distance spacers are joined with M16 bolts of strength class 5.8, while the distance spacer and the posts are joined with M10 bolts of the same strength class. The later are supposed to fail in the early stages of the crash so that the guardrail is released from the post instead of being pulled down with it.
Fig. 3. Elements of the road barrier.
Z. Ren, M. Vesenjak / Engineering Failure Analysis 12 (2005) 963–973
967
5. Description of the computational model A 34 m long segment of the road safety barrier was found to be sufficient for conducted impact analyses. The guardrail was modelled as a continuous element under presumption that the guardrail splices were not the weakest link, according to Engstrand [5]. The model comprised also 9 posts separated by the distance of 4 m. All component parts were modelled with the 4-node Belytschko-Tsay shell elements with three integration points through the shell thickness [6,7]. Thicknesses of model parts are the following: guardrail 3 mm and distance spacer and post 4 mm. The impacting vehicle, type Metro (publicly accessible on web page: http://www.ncac.gwu.edu), was modelled mostly by shell elements and its weight was tuned to prescribed 900 kg (Fig. 4) [8]. 5.1. Material model All parts of the road restraint system are made of the construction steel S 235. Standard tensile tests of different thickness specimens were carried out according to the standard DIN 50 115 to obtain realistic material properties of analyzed road restraint elements in terms of their YoungÕs modulus, yield stress, elasto-plastic relationship and ultimate strength. Obtained material parameters were used to define isotropic bilinear elasto-plastic material model with kinematic hardening in subsequent computational analyses. Previously experimentally determined dynamic material response parameters were considered in all computational analyses. It was presumed that the bolts and welds are rigid. The level of effective plastic strain was used as a failure condition in the model. The maximum effective strain was set to 0.28 which corresponds to 28% ductility of S 235 steel. When the effective plastic strain reached that value, the element load carrying capability (stress) was reduced to zero, effectively removing the element from the model. 5.2. Boundary conditions and contact description The posts were fixed with viscoplastic springs (Fig. 5) in order to simulate the soil influence. Linear springs were attached at both ends of the guardrail (Fig. 5) to simulate its continuation. All surfaces of the model were defined as one contact group, thus, effectively accounting for multiple selfcontacting regimes during computational impact analyses. The static and dynamic friction coefficient between all parts was set to 0.1 and 0.05, respectively.
Fig. 4. Impacting vehicle.
968
Z. Ren, M. Vesenjak / Engineering Failure Analysis 12 (2005) 963–973
Fig. 5. Boundary conditions.
5.3. Dynamic analysis properties The vehicle was prescribed to have the initial velocity of 100 km/h at an impact angle of 20 with regard to the guardrail (Fig. 5). The analysis time interval was set to 350 ms, with results output required every 1 ms. The time step for explicit transient dynamic analysis was set to 1.4 ls by the used code LS-Dyna PC Version 970 with regard to the lowest resonant frequency of the structure. The analyses run 85– 95 h on a PowerPC with 2 AMD 2000+ MHz processors and 2 GB RAM, depending on the complexity of the individual model.
6. Analysis of computational results The results of computational simulations are shown in Figs. 6–9. Fig. 6 shows the deformation steps of the safety barrier and appropriate containment of the impacting vehicle. The computed contact length between the barrier and the vehicle was 10.8 m. Computational simulations have shown that extreme decelerations of the vehicle mass centre occur during initial vehicle impact at the guardrail and later when the vehicle hits the post. ASI index was well bellow the safety limit and was the highest when the front wheel hit the first post (max ASI = 0.66), see Fig. 13. Simulations have also clearly proved that appropriate distance spacer design, through its controlled deformation, significantly reduces the decelerations experienced by vehicle occupants in the initial phase of an impact. The hexagonal distance spacer deforms in a controllable manner until the guardrail is pushed to the post (Fig. 7) and absorbs the highest possible amount of crash energy. From Fig. 8, it is also obvious that a deceleration peak occurs when the front wheel hits the post. With a guidance profile those high decelerations can be avoided. The simulations have also shown that the weak connection between the distance spacer and the post failed as designed. The connection failed just before the vehicle hit the first post (Fig. 9). The guardrail was released and consequently contained the vehicle on the road. The working width (maximum lateral dynamic deformation) of the simulated safety barrier was equal to 1.2 m, which corresponds to working width class W4 according to EN-1317-2.
Z. Ren, M. Vesenjak / Engineering Failure Analysis 12 (2005) 963–973
969
Fig. 6. Computational simulation of the vehicle impact.
7. Full-scale crash test ¨ V Automotive test facility in GerThe new road safety barrier was tested in a full scale crash test at TU many according to the EN 1317 standard. The crash test was carried with the FIAT Uno as the impacting vehicle with the same initial impact conditions as those used in the simulations (Fig. 10). The road restraint system performed well and was successful in redirecting the vehicles back on the road without collapsing (Figs. 11 and 12). The measured maximum ASI index was equal to 0.63 (Fig. 13), the working width was 0.95 m (class W3 according to EN-1317-2) and the length of the contact between the impacting vehicle and the barrier was equal to 12 m.
970
Z. Ren, M. Vesenjak / Engineering Failure Analysis 12 (2005) 963–973
Fig. 7. Distance spacer deformation.
Fig. 8. Wheel impact at the post.
Fig. 9. Weak connection between the distance spacer and the post.
Z. Ren, M. Vesenjak / Engineering Failure Analysis 12 (2005) 963–973
Fig. 10. Crash test vehicle (FIAT Uno) (a) before and (b) after the impact.
Fig. 11. Test vehicle impact at the road restraint system.
Fig. 12. Deformation of the road restraint system.
971
972
Z. Ren, M. Vesenjak / Engineering Failure Analysis 12 (2005) 963–973 1.00 0.90 computational results 0.80
experimental results
0.70
ASI
0.60 0.50 0.40 0.30 0.20 0.10 0.00 0.00
0.05
0.10
0.15
0.20
0.25
0.30
0.35
Time [s]
Fig. 13. Computational and experimental results comparison.
8. Comparison of the computational and experimental results From comparison of computational (Section 6) and experimental (Section 7) results it can be observed that the results compare very well in relation to the working width of the system (computed 1.2 m; measured 0.95 m) and contact length (computed 10.8 m; measured 12 m). The differences can be attributed to parameters used to describe material dynamic behaviour in computations, which were obviously underestimating the stiffness increase of the material under dynamic impact loading. Computational and experimental ASI index histories are shown in Fig. 13. ASI index for the tested system does not exceed the limiting value in any case. The highest values (computational 0.66; measured 0.63), are observed when the front wheel hits the first post. 9. Conclusions Computational nonlinear explicit dynamic analysis was employed for evaluation of the new road safety barrier behaviour under test vehicle impact conditions. Computational simulation predictions have been compared with the results of the full-scale crash test. The used computational model provided comparable results to experimental measurements and can thus be used for computational evaluation of other road safety barriers in order to avoid numerous expensive full-scale crash tests. The tests have also shown that the new safety barrier assures controllable crash energy absorption which in turn increases the safety of vehicle occupants. References [1] Washington State Department of Transportation. Roadside safety-design manual, Washington; 2003. [2] Road Safety 2010 – A Framework for Saving 2000 Lives by the year 2010. Roads and Traffic Authory, New South Wales.
Z. Ren, M. Vesenjak / Engineering Failure Analysis 12 (2005) 963–973 [3] [4] [5] [6] [7] [8]
973
European Committee for Standardization. European Standard EN 1317-1, EN 1317-2, Road Restraint Systems; 1998. Vesenjak M, Ren Z. Computer-aided design of a road restraint barrier. J Mech Eng 2003;49(10):509–19. Engstrand KE. Improvements to the weak – post W – Beam Guardrail. Worcester: Worcester Polytechnic Institute; 2000. Livermore Software Technology Corporation, LS-DYNA keyword userÕs manual; 2001. Livermore Software Technology Corporation, LS-DYNA theoretical manual; 1998. Vesenjak M, Ren Z. Dynamic analysis of a road restraint systemÕs deformation resulting from a vehicle impact. J Mech Eng 2003;49(12):586–92.