Proposal for a dynamic rollover protective system test S Richardson*#, R H Grzebieta* and G Rechnitzer° *Monash University Department of Civil Engineering, Australia °Monash University Accident Research Centre (MUARC), Australia # Land Engineering Agency (Australian Army) Abstract: Currently effective Rollover Protective Structure standards exist for Buses, Earthmoving Equipment and Agricultural Tractors. Construction guides for racing cars have also evolved. There are no well-recognised Rollover Protective System tests, standards or design guides for either Military or Civilian (4 Wheel Drive Vehicles) 4 × 4’s. This paper presents a range of test protocols, procedures and demonstration methods that have been developed and used to evaluate and test Rollover Protective Structures for Military 4 × 4’s. A dynamic test method developed by Exponent, Inc. is also examined. Based on the identified weaknesses of the above and the desire to develop a reproducible test method that simulates real world rollover crashes, a dynamic Rollover Protective System test is proposed. The proposed Test will enable evaluation of active safety systems, occupant restraints and structural performance via injury criteria developed for Anthropometric Test Dummies (ATD). A feasibility study of the proposed test concept will be carried out via MADYMO modelling and 1/4 scale structure modelling prior to physical build of the test facility. Australian Army experience of rollover BACKGROUND The majority of rollovers within the Australian Army occur in the most numerous vehicle, the Perentie 4 × 4 soft-top variants. Based on Land Engineering Agency data the Australian Defence Force has had 125 rollovers resulting in 12 deaths 45 serious injuries and 67 minor injuries in the period Jan 1988 to Feb 1996. This was a rate of 1.3 rollovers per month. The Safe Carriage of Troops [4] (SCOT) study identified that post 1996 the rate of rollovers dropped due to the implementation of Defence Instruction – General (DI-G) 19-6. However, the main impact of 196 was to restrict use and carriage of personnel in vehicles like the Perentie 4 × 4 variants. Examination of the SCOT data indicates that DI-G 19-6 significantly reduced the km travelled. Nevertheless, the rollover rate per 10,000 km pre and post DI-G 19-6 remained similar. The Australian Army has instigated several projects to investigate and implement improvements to the crashworthiness of its existing and future fleets of vehicles. The following sections of this paper summarise the work carried out to date by the Australian Army (Land Engineering Agency) to develop Rollover Protective Structures for Military 4 × 4’s (vehicles with a fully laden mass < 5,000 kg). The way in which soldiers have been transported in military vehicles has not changed significantly in the past 80 years. Typically soldiers are expected to sit on side facing wooden bench seats without restraints and the vehicles have no effective Rollover Protective Structure. The predominant crash mode for all classes of off-road Military vehicles is rollover [1]. Rollover is also a significant problem for Civilian 4 × 4’s with rollover rates of up to 5 times that of the average passenger car, on roads [2]. The problem of rollover is magnified further when vehicles are operated in a non-urban environment. Rechnitzer [3] et al. reported in an Australian based study on rollover that; “rollover crashes are a common cause of occupant injury especially on non-urban roads. Their importance increases with injury severity: they constitute 19% of occupant fatalities in Australia. This percentage rises to 44% in rural Western Australia and 54% in rural Northern Territory.” Corresponding Author: Shane Richardson, DVExperts Pty Ltd, 229 Canterbury Road Canterbury, Vic 3126, Australia Tel +61 3 9880 7399 Fax +61 3 9880 7725 Email srichardson@dvexperts.net © Woodhead Publishing Ltd 0215 Project PERENTIE In 1986 the Australian Army awarded a contract to Jaguar Rover Australia for Project PERENTIE to procure 133 IJCrash 2003 Vol. 8 No. 2 pp. 133–141 S Richardson, R H Grzebieta and G Rechnitzer The rollover crash investigations identified that the structural hoop was forming a plastic hinge at its base. As a result soldiers were being exposed to severe and fatal head and neck injuries. The location of the upper seat belt sash anchor on the forward structural hoop also caused injury and fatalities by crushing the occupant’s chest between the seat belt and seat. Ludovici [8] produced a requirement using three different load cases, all based on rigidly fixing the upper seat belt sash anchor and providing a Rollover Protective Structure for the front two seat occupants. The load cases were: replacement vehicles for the Series 3 Land Rover. The two basic vehicles procured were a Perentie 4 × 4 and 6 × 6, with 1 tonne and 2 tonne payload capacities respectively. The Perentie 4 × 4 is a militarised derivative of the Land Rover 110. A range of variants were produced, the most numerous being a soft-top 4 × 4 (Figure 1). The concept document for the Perentie [5] identified the requirement for a vehicle Rollover Protective Structure but this was not in the final specification. It is assumed that this safety requirement was removed as part of a capability trade off. The trade off made was to increase the length of the front and rear suspension coil springs of the standard Land Rover 110 by 44 mm and 106 mm [6] respectively for the Perentie 4 × 4. Quite possibly these two trade offs, removal of the Rollover Protective Structure and lengthening of coil springs adversely affected the crashworthiness and handling of the Perentie 4 × 4. The soft-top 4 × 4 was initially fitted with two structural hoops at the front and rear of the cargo area to support the canvas canopy and allow camouflage poles and nets to be carried on the roof. The front hoop also provided a sash anchor for the driver and front seat passenger lap sash seat belts (Fig. 1). No restraints were fitted for occupants in the rear cargo area. In 1989 the Perentie 4 × 4’s was introduced into service and rollover crashes [7] were reported. After 30 rollovers, 3 deaths, 9 serious injuries and 23 minor injuries due to rollover; the Perentie Project Office began an investigation of the crashes. Two issues were identified and evaluated. 1. Fx = –6.2 kN, Fy = –28.2 kN and Fz = –37.7 kN applied to either upper structure corner. 2. Fx = –25.1 kN, Fy = –15.7 kN and Fz = –12.6 kN applied to either upper structure corner. 3. Fx = –42.4 kN applied centrally to the horizontal upper part of the structure. Various Rollover Protective Structure designs were proposed by Army Agencies. Monash University, Civil Engineering conducted a Finite Element Analysis (FEA) of alternative designs [9]. Monash reported that the preferred Army Part 4 design (Fig. 2) would sustain all loads except front to rear loading. The Monash report stated “at best the structure may sustain 62% of the Fx load”. The following recommendations proposed by Monash were not adopted. 1. Brace the whole rear region of the tray as shown in Figure 3 as an inexpensive way of greatly increasing the rollover safety of the vehicle. 2. Increase the cross-sectional dimensions of the roll bar so as to significantly increase the amount of energy absorbed during the first impact. 1. A Rollover Protective Structure to protect the driver and front seat passenger. 2. The handling performance of the Perentie 4 × 4 in comparison to the Series 3 Landrover (the vehicle that the Perentie 4 × 4 replaced), Nissan Patrol and Toyota Land Cruiser. Note: Seat belt sash guide anchored on the forward structural hoop. Figure 1 1988 Soft-top Perentie 4 × 4 Utility: only the forward structural hoop is shown at the front of the cargo area. IJCrash 2003 Vol. 8 No. 2 134 0215 © Woodhead Publishing Ltd Proposal for a dynamic rollover protective system test Following the development of the Rollover Protective Structures Ludovici [10] produced a discussion paper identifying that the Part 4 and RFSV design was weak to reverse loading and that passengers in the rear of the vehicle were not adequately protected. The Ludovici paper proposed: Rearward 40 kN 7 kN 1. A test method similar to 86/295/EEC rollover protection directive for construction equipment. 2. The test forces as calculated (EEC standard) with an additional consideration given for forward movement of the vehicle. 3. The vertical test force as specified in AS1636 be adopted. 4. The seatbelt anchor point rigidity be combined with a suitable Survival Space. 40 kN Forward z y x Figure 2 The Part 4 ROPS design. No action was taken to implement the proposals of the Ludovici discussion paper. Forward Project BUSHRANGER Project BUSHRANGER was initiated in 1992 with the aim of providing a mix of unarmoured and armoured vehicles to motorise selected Infantry Battalions within the Australian Army. There are three phases to the Project. Phase 1 (1994–98) was to provide unarmoured vehicles for interim motorisation while the armoured vehicle was being selected. The Phase 1 vehicles are based on the Perentie and were procured using an option within the Project Perentie contract. Phase 2 (1998–99) was a competitive evaluation of alternative armoured vehicles. Phase 3 (2000–05) is the production and introduction into service of the selected armoured vehicle. The specification for the Phase 1 vehicles included a Rollover Protective Structure and seat restraints that would protect all the crew in a forward or rollover crash. The Structural specification was based on previous Project Perentie work and defined three load cases and a Survival Space, which could not be infringed upon. The load cases for the BUSHRANGER Phase 1 4 × 4 [11] were: Rearward Figure 3 Monash proposed ROPS. The Part 4 design (Figure 2) was accepted by the Perentie Project Office and fitted to the Utility and FFR variants of the Perentie 4 × 4. The Regional Forces Surveillance Vehicle (RFSV), another soft-top 4 × 4 variant, was specifically built for a reconnaissance role with a three-person crew. A different Rollover Protective Structure was developed and fitted to the RFSV based on its external profile. No testing was carried out to prove either the Part 4 or the RFSV Rollover Protective Structure or to validate the FEA. Although the Perentie 4 × 4s continued to rollover in the field the Perentie Project Office observed that the death and serious injury rate due to Perentie rollover had diminished and ceased investigations into the handling of the Perentie 4 × 4a. The Project Office concluding that through a combination of improved driver-training, education, licensing controls and the addition of an effective Rollover Protective Structure, the rollover problem had been solved. This conclusion was not validated. 1. 13 kJ to the front upper corner of the Structure. 2. 10 kJ to the rear upper corner of the Structure. 3. 72 kN and 5 kJ applied to the forward upper horizontal part of the Structure. Impacts 1 and 2 were conducted on the same side of the vehicle. The size and shape of the Survival Space was modified during testing to include a desirable Survival Space based on SAE J154 [12] occupant shapes and a smaller essential Survival Space. The Bushranger Phase 1 4 × 4’s were certified using the smaller essential Survival Space. The method to input the defined energy levels was not defined. a The Australian Army (Land Engineering Agency) has conducted a range of investigations into the handling of the Perentie 4 × 4 variants, which concluded in Dec 2001. A modification package, which significantly improves the handling of all the Perentie 4 × 4 variants, has been developed. The modification package involves fitting an anti-rollbar to the front axle and lowering the chassis attachment points of the rear axle lower trailing control arms. The modification package is currently being considered for implementation within current Defence resource limitations. © Woodhead Publishing Ltd 0215 Project TRANSAFE During the period 1992–94 there were still a number of crashes where soldiers were killed or injured as a result of 135 IJCrash 2003 Vol. 8 No. 2 S Richardson, R H Grzebieta and G Rechnitzer that it was relatively simple to design an open frame structure to withstand the loads. However, transferring the loads into the vehicle proved a more difficult and frustrating problem to solve because the vehicle collapsed around the tubular structure or the structure detached from the vehicle during testing (Fig. 5). In parallel with the physical testing, Monash University, Department of Civil Engineering [15] modelled the structure (Fig. 6). Good correlation was achieved between the tests and the Finite Element Analysis of the Rollover Protective Structure with respect to deformation. During the development of the TRANSAFE Rollover Protective Structure performance requirement, examples of rolled over Perentie 4 × 4 were continually sought out and, where possible, examined and the crash investigated. A specific case was the rollover of a RFSV that had been modified and fitted with a RFSV TRANSAFE structure and occupant restraints. The purpose of the modification was to gain feedback and input as to the effects on operations of the various safety systems installed. One week after modification and while returning from a patrol, the RFSV rolled over, after the driver lost control and slewed the vehicle into a ditch. At the location where the vehicle left the road (Fig. 7) it was travelling at approximately 80 km/h and slewed 150° to the right (pointing backwards). The vehicle impacted the ground at least 4 times and is estimated to have completely rotated 5 times (20 1/4 turns). All 3 occupants survived; the driver had a 5 cm long cut to the forehead; the front left occupant had no injuries and the rear seat occupant was knocked unconscious. Figure 8 illustrates the RFSV after the rollover. The crash was investigated and the report [16] concluded: being carried in the rear of military vehicles. As a result the Australian Army initiated Project TRANSAFE to examine Transportation Safety with respect to unarmoured, wheeled military vehicles. One of the conclusions of an initial engineering assessment of the options [13] for the Australian Army was “a well-designed Rollover Protective Structure will provide effective Survival Space for the occupants and if used in combination with seat restraints will improve occupant safety”. An investigation to retrofit the Perentie 4 × 4 proceeded. However, the available standards and test methodologies were considered inadequate or inappropriate. Of the available standards either the magnitude of the applied load appeared too small or the direction of principle force was incorrect, based on the investigated Perentie 4 × 4 rollover crashes. A TRANSAFE Rollover Protection Structure specification was developed to protect all vehicle occupants. The loading direction from the Bushranger specification revised loading requirements, and the Survival Space located in seating positions and derived from SAE J154 occupant shapes, were used. The load cases were: 1. 13 kJ to the front upper corner of the Structure. 2. 10 kJ to the rear upper corner of the Structure. 3. 5 kJ applied to the forward upper horizontal part of the Structure. Load Cases 1 and 2 were conducted on the same side of the vehicle. A impact pendulum was used to apply the energy, based on the pendulum requirements within Australian Design Rule 59 [14]. A conceptual Rollover Protective Structure was developed (Figure 4) utilising the existing Part 4 structure and the rear structural hoop. The rear structural hoop was braced diagonally in the same manner as the part 4 design. The Part 4 and rear structural hoop were connected longitudinally and braced by two rings. The rings provide an excellent structural component and also permit military considerations such as fitting a weapon mount or allow the soldiers to stand in the rear of the vehicle and dominate a position. 1. There was nothing mechanically wrong with the vehicle. 2. The TRANSAFE modification had not caused the accident. 3. The accident was a reverse rollover. The vehicle was upside down and travelling backwards when it first hit the ground. If the occupants had been in a conventional RFSV, the rear seat occupant would have been killed and the two front seat occupants would have been seriously injured or most probably killed. The analysis of the above crash and others indicated an inconsistency with the specification. It was decided to alter the order of loading to; upper front, upper horizontal and upper rear with the upper rear loading on the opposite side to the upper front. The loading to the upper horizontal was also changed from an energy to a force criterion. The determination of both force and energy criteria were also altered so that values were determined using vehicle geometry and combat laden mass. The load cases for the TRANSAFE Perentie 4 × 4 were: Forward Figure 4 Conceptual TRANSAFE Perentie 4 × 4 ROPS. To examine options of tube size and attachment methods and to determine if a retrofit solution was practical and economically feasible, several Rollover Protective Structures were built on vehicles and tested. The testing illustrated IJCrash 2003 Vol. 8 No. 2 1. 13 kJ to the front upper corner of the Structure. 2. 72 kN applied to the forward upper horizontal part of the Structure 3. 10 kJ to the rear upper corner of the Structure. 136 0215 © Woodhead Publishing Ltd Proposal for a dynamic rollover protective system test Figure 5 ROPS detaching during testing. Forward Y X Forward Figure 6 Monash Finite Element Analysis model of the TRANSAFE ROPS. Figure 8 Rolled over RFSV. Loading Con’d 2 69° Loading Con’d 3 49° 69° 49° 7° Loading Con’d 1 Figure 7 View back up the road from the point the RFSV left the road. Impacts 1 and 3 are conducted on opposite sides of the vehicle, as illustrated in Figure 9. The weakness with the TRANSAFE Performance Specification is that it is a structural test used in combination with Survival Space(s) to evaluate a Rollover Protective System. No consideration is given to occupant kinematics, injury criterion or the performance of the restraints. © Woodhead Publishing Ltd 0215 Figure 9 Orientation of ROPS Loading Conditions. TRANSAFE dynamic demonstration In order to provide some assurance of the dynamic performance of designs developed for the TRANSAFE Performance Specification and to provide effective visual images to convince soldiers to use the developed systems, a dynamic demonstration was created. 137 IJCrash 2003 Vol. 8 No. 2 S Richardson, R H Grzebieta and G Rechnitzer The dynamic demonstration involved dropping a vehicle onto a dirt road, suspended 1 m off the ground behind a large Tow truck travelling at 50 km/h. The vehicle dropped was orientated with 30° of forward rotation, 30° longitudinal superelevation and offset 45° to the direction of motion. Figure 10 shows two photographs illustrating the dynamic demonstration. The intent of the dynamic demonstration was to control the position of the vehicle for the first rollover impact. Various experiments were conducted to refine the methodology with the variables being; speed, drop height, longitudinal superelevation angle, roll angle, inclination to the direction of motion and surfaces that the vehicle is dropped onto. During the experiments, changes in the variables resulted in the vehicle rotating 121/4 turns, breaking the back of the vehicle or disintegration of the Rollover Protective System. The objective was to create equivalent deformation to the structure as in the Performance Specification test, which in turn was based on investigated rollover crashes. The weakness with the drop method is that only one vehicle type has been used (Perentie 4 × 4) and repeatability has not been validated. No Athopromorphic Test Dummies (ATD’s) have been used and given the pre-drop orientation of the vehicle, systems designed to detect rollover and initiate pyrotechnic restraint devices can not be evaluated. DYNAMIC ROLLOVER PROTECTION SYSTEMS TEST REQUIREMENTS In order to explore and evaluate the Rollover Protective System a dynamic test methodology and alternative test protocol is needed, much in the same way that the Frontal, Offset and Side impact tests are conducted by consumer based New Car Assessment Programs (NCAP). Thus the Rollover Protective Systems, as a whole, could be evaluated rather than individual components, i.e. effectively reflecting a “standard and reproducible” real world rollover crash. One consistent and justifiable criticism of dynamic rollover tests is lack of repeatability. It was considered that part of the problem with alternative dynamic rollover tests, such as ramp induced rollovers [17,18] or the SAE J2114 [19] is controlling the vehicle so that a consistent first impact is achieved. A dynamic Rollover Protection System test should simulate the typical observed rollover, which involves the vehicle yawing across the road, leaving the road in a sideways configuration, tripping and rolling over (Fig. 11). It is considered that the spiral roll test procedure being developed in Europe [20] dose not serve this purpose. The issues identified that are important in the development of a dynamic rollover protection systems test procedure are: Figure 10 TRANSAFE dynamic demonstration. Figure 11 Typical rollover. IJCrash 2003 Vol. 8 No. 2 138 0215 © Woodhead Publishing Ltd Proposal for a dynamic rollover protective system test 1. 2. 3. 4. 5. 6. and FMVSS 201 assesses interior head impact protection. There is however no design rule or rating system that assesses the effectiveness of occupant protection systems in the vehicle such as seat belts, curtains, etc in the case of a rollover crash. There is considerable debate regarding roof crush strength, the adequacy of FMVSS 216 and its relationship to occupant injury. [24,25,26,27,28]. The key to this argument is that the whole system, i.e. roof and seat belt restraints, belt pre-tensioning, side curtain or side air bag, must be assessed to provide adequate protection to occupants in such crashes. A dynamic rollover test clearly needs to addresses these issues. Rollover metric Ejection Roof strength Injury criterion of occupants Correlation to real world crashes Repeatability Rollover metric Wilber [21] presented Table 1, which clearly shows that rollovers 41/4 turns or more are the most injurious and fatal. Table 1 Injuries v’s 1/4 turns 1 /4 turns Injuries AISb 1-2 Injuries AIS 3-6 14.4% 34.3% 9.0% 42.4% 11.3% 38.4% 7.7% 42.5% 6.9% 16.3% 7.3% 69.5% One Two Three ≥Four b Injury criteria for occupants Single vehicle lateral rollovers Debate has also ensued regarding which dummy is most appropriate to use for assessing injuries that may be inflicted in a rollover crash. Hybrid III dummies were used in the Malibu tests [25,26]. There is considerable criticism regarding the neck loads measured and their relation to real world rollover neck injuries. Autoliv in Sweden has developed a Rollover Dummy with a biomechanical spine that may better represent the human spine. It has also been suggested that the Thor dummy displays better biomechanical response. Hence, research work still needs to be carried out to assess if a simple combination such as Hybrid III with a suitable neck developed for rollover, would provide an injury envelope that encompasses the biomechanical response of the more sophisticated and expensive dummies. Work would also need to be carried out reviewing real world rollover crashes so as to determine which injury criteria would be relevant to the dynamic test protocol outlined below, i.e. HIC, neck and possibly chest. AIS Abbreviated Injury Scale One of the fundamental flaws that are attributed to analysis of rollovers is that rollovers can be compared using a Delta V metric. In a frontal, offset, side and or rear collision, Delta V is a good predictor of occupant injuries. However, in the case of rollover the analysis by Wilber suggests that 1/4 turns can be used. The sliding velocity required to produce 41/4 turns [22] is based on an energy calculation using a trapezoidal geometry profile (rolling cross section of the vehicle), the second moment of mass inertia [23] about the roll axis, vehicle mass and the centre of gravity location. (Eg. the 41/4 turn sliding velocity for a 1990 Mitsubishi Pajero is 38.8 km/h). The aim for the Rollover Protection Systems test is to impose between 4 to 61/4 turns. In other words at a minimum at least one complete rotation of the vehicle. Correlation to real world crashes The test protocol needs to be compared with actual real world crashes to determine the correlation between observed injuries, occupant kinematics and vehicle damage. Ejection Repeatability Occupant full or partial ejection is a major cause of serious injury in rollover crashes [24]. It is preferable for the occupants to be restrained during a rollover crash. The question arises as to what constitutes an appropriate restraint. Obviously a restraint system that allows full or partial ejection is inadequate. Pretensioned seat belts and side airbag curtains can be triggered to provide good containment. Hence a dynamic procedure that tests the triggering (including the sensors) of such restraint systems and assesses if any part of the occupant’s body extends outside the structure is considered essential. In summary the Rollover Protection System test should be capable of assessing ejection, the adequacy of restraint systems and the triggering mechanisms. The main purpose of the test protocol is to provide a repeatable test. This will be achieved utilising the repeatability performance parameters from barrier and side impact tests. Proposed Dynamic Rollover Test Protocol Figure 12 shows the Controlled Rollover Impact System (CRIS) developed by Exponent, Inc in the USA. The CRIS rig is towed on the back of a truck and the vehicle is continuously rotated prior to release. The ATDs are tethered prior to release. The test vehicle is released using explosive bolts. The release of the tethers and vehicle is computer controlled and provides a repeatable output. The main weakness with CRIS is that the vehicle is continually rotating prior to being released and this requires that the ATDs be tethered. The continual rotation also precludes test and evaluation of active safety systems such as side curtains or seatbelt pre-tensioners. Roof strength US design rule Federal Motor Vehicle Safety Standard (FMVSS) 216 for assesses the strength of a vehicle’s roof © Woodhead Publishing Ltd 0215 139 IJCrash 2003 Vol. 8 No. 2 S Richardson, R H Grzebieta and G Rechnitzer Figure 12 Rollover CRIS test structure developed by Exponent. This Rollover Protection System test would enable an assessment of a vehicle’s occupant protection systems via injury criteria of ATD’s. This systems test would include the evaluation of the rollover trigger sensor systems as well as on-board protection devices and vehicle structure as a system test. Monash are currently developing MADYMO models of the proposed Rollover Protection System test and intend to develop a 1/4 scale structure to validate repeatability prior to the construction of the a full scale structure. This is a major but worthwhile research initiative and international co-operation is being sought to develop the test procedures. However, CRIS produces repeatable tests and is a significant step in the evolution of a dynamic Rollover Protection System test. To provide a repeatable test technique for evaluating a roof-to-ground impact, sensor triggering and occupant protection in a vehicle rollover event, it is essential that a translating and rotating vehicle drop system be developed. The system must also be compact enough so that it can be located in a crash test facility. The system would be constructed such that it holds a vehicle at a pre-set height and selectable roll, pitch, and yaw attitudes similar to the CRIS system (Figure 11). The restraining structure needs to be made adjustable so that the vehicle can be oriented anywhere in a three dimensional volume (within reasonable limits). The structure would be propelled forward (towed if in a crash test facility) at a constant preset speed such that the test vehicle leads the rig. At a predetermined point the structure is decelerated at around 0.5 g’s (or at a deceleration rate simulating a pre-rollover yaw). This deceleration positions the dummies as in an actual rollover event. A flywheel on the rig would then be engaged to initiate vehicle rotation and milliseconds later the car is released at the 41/4 turn sliding speed. Deceleration, rotation and release would need to be synchronised to ensure proper vehicle orientation upon contact with the ground. High-speed cameras suspended from the test fixture and inside the vehicle would provide detailed information regarding firing of on board protection systems and occupant-roof-ground interactions. Such views are not available in less-controlled rollover testing techniques, such as dolly rollovers. This Rollover Protection System test would be different from the CRIS system in that: ACKNOWLEDGEMENTS The development of the TRANSAFE performance specification and dynamic demonstration would not have been possible without the assistance of colleagues at the Land Engineering Agency, specifically R. Hosemans, N. Dyson, A. Sims, R Andriuzzi, the staff of both the Prototype Engineering Centre and Accredited Test Services. REFERENCES 1. 1999 SAE TOPTEC on Military Vehicle Safety, San Francisco, Jun 1999. 2. SNYDER, R G, MCDOLE, T L, LADD W M and MINAHAN, D J. On-road crash experience of utility vehicles University of Michigan, UM-HSRI-80-14. 3. RECHNITZER, G, LANE, J and SCOTT, G. Rollover Crash Study – Vehicle Design and Occupant Injuries, 15th ESV, 96-S5-O-10. 4. RECHNITZER, G, HAWORTH, N, DIAMANTOPOULOU, K, GRIFFITHS, P, STEVENS, M, CLAYTON, A, WHEELER, W, KINGHORN, R, POLKINGHORNE, G and DUTTON, M. Safe Carriage of Personnel in General Service Vehicles Study, Aug 1998. This report was prepared for the Australian Defence Force, Land 121, Project Overlander. 5. Australian Army, Army Staff Requirement 63.11, Army’s Range of light Field Vehicles issued 13 Feb 1981. 6. SIMMONS, K. (MAJ.), Proposal to investigate Handling Improvements to the Perentie 4 × 4 Family of Vehicles, • • • • it is propelled with the test vehicle forward of the rig the test vehicle is held so that it is not continuously rotating deceleration is used to position the dummies an inertia fly-wheel is used to initially roll the test vehicle over • the rig will be compact enough to fit into a crash test laboratory IJCrash 2003 Vol. 8 No. 2 140 0215 © Woodhead Publishing Ltd Proposal for a dynamic rollover protective system test 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. Cranfield University, School of Defence Management, MA Dissertation 1994. Maintenance Engineering Agency, File 2320-80-1, Minute Landrover 110 (Perentie), dated 23 Oct 89. Engineering Development Establishment, Mobility Engineering Division, Land Rover 110 4 × 4 – Rollover Protection Finite Element Analysis of Rollbar Structures. GRZEBIETA, R. Rollover Analysis of a 4 × 4 Landrover 110 Rollbar, Monash University, 1989. LUDOVICI, D. Discussion Paper – Roll Over Protection Standard, Rollover Protection and Roll stability Analysis of Landrover 110 4 × 4, Technical Working Group, Engineering Development Establishment, Sep 1990. ARMY (AUST) 6835, Project Bushranger Phase 1, Annex C. Operator Space Envelope Dimensions for Off-Road Machines, SAE J 154. KILLINGWORTH, W and RICHARDSON, S. Project TRANSAFE 94 Phase 1, Engineering Development Establishment, dated 20 Jun 94. Australian Design rule 59; Omnibuses Rollover Protective Structures. RICHARDSON, S and GRZEBIETA, R. Roll Over Protective Structures for the Safe Carriage of Troops in the Rear of Military Vehicles, Monash University Industry Geomechanics and Structures Symposium 96. RICHARDSON, S. 2300-Y1-54, ATEA Minute “Note for file” dated 15 Sep 95. WILSON, R and GANNON, R. Rollover Testing, SAE 720495. WECH, L and OSTERMANN, B. The safety of convertibles in realistic rollover crashes, 15th ESV, 96-S5-O-02. Dolly Rollover, Recommended Test Procedure-Society of Automotive Engineers J2114. DIGGES, K and KLISCH, S. Analysis of the factors which © Woodhead Publishing Ltd 0215 21. 22. 23. 24. 25. 26. 27. 28. 141 influence rollover crash severity, 13th International Technical Conference on Experimental Safety Vehicles, NHTSA, 1991. WILBER, V H. American Automobile Manufactures Association, Vehicle Rollover Prevention a balanced approach to a complex problem, 15th ESV, 96-S5-O-07, 1996. RICHARDSON, S, GRZEBIETA, R H and BELLION, P. Proposed 41/4 turn metric to simulate rollover crashes, 17th ESV 01S6-W-99, 2001. http://www nrd.nhtsa.dot.gov/vrtc/ca/nhtsa_ inertia_database_metric.xls HENDERSON, M and PAINE, M. Passenger Car Roof Crush Strength Requirements, report No. CR176, Federal Office of Road Safety, Department of Transport and Regional Development, Australia. ORLOWSKI, K, BUNDORF, R T and MOFFAT, E A. Rollover crash test: the influence of roof strength on injury mechanics, SAE 851734, in Proceedings 29th Stapp Car Crash Conference, Society of Automotive Engineers 1985. BAHLING, G S, BUNDORF, R T, KASPZYK, G S, MOFFAT, E A, ORLOWSKI, K F and STOCKE, J E. Rollover and drop tests: The influence of roof strength on injury mechanisms using belted dummies, SAE 902314, in Proceedings, 34th Stapp Car Crash Conference, Society of Automotive Engineers, 1990. FRIEDMAN, D and FRIEDMAN, K D. Roof Collapse and the risk of Sever Head and Neck Injury, in Proceedings, 13th International Technical Conference on Experimental Safety Vehicles, 1991. SYSON, S R. Occupant to roof contact: roll-overs and drop tests, SAE 950654, in advances in Occupant Protection Technologies for the mid Nineties, SAE SP 1077, Society of Automotive Engineers, 1995. IJCrash 2003 Vol. 8 No. 2 IJCrash 2003 Vol. 8 No. 2 142 © Woodhead Publishing Ltd