The vehicle model was an FEA representation of the bumper, bonnet and windscreen of a mid-size, 1.8 litre car. The vehicle model was created by measurement of the actual vehicle involved in the incident using a three-dimensional co-ordinate measuring rig. Profiles obtained were compared to manufacturer data.
Pedestrian orientation with respect to the vehicle were varied across three positions:
facing the vehicle, side on the vehicle and facing away from the vehicle at a 45 degree angle (Refer Figure 3.8). Other orientations were determined to produce unrepresentative leg injuries. The pedestrian model used was the MADYMO 5th percentile female multibody human model, Version 6.01. Other simulation parameters can be seen in Table 3.4. A discussion of the literature values and ranges referred to can be found in Chapter 6.
Parameter Range Comment Coefficient of friction
between vehicle and pedestrian
0.45 Within range of values
reported in literature
Coefficient of friction between pedestrian and ground
0.55 for pedestrian on ground, 0.7 for shoe contact on ground
Value indicated to be within literature values and those determined by author Vehicle speed at impact 5.56 to 9.72 ms-1 varied in
1.39 ms-1 increments
Range determined pedestrian injuries and damage to vehicle cowl using guidelines from Happer et al (2000)
Vehicle acceleration -8.5, -4.0 and 0.0 ms-2 Maximum vehicle deceleration (determined
Stiffness of vehicle bonnet top
300 Nmm-1 As per literature.
Pedestrian head stiffness 2500 Nmm-1 As per literature for anterior-posterior loading.
Stiffness of road 40 kNmm-1 Middle of range specified by Chadbourn et al (1997)
Table 3.4 Parameters for Simulation of Collision Involving a Typical Vehicle
The contact model chosen for modelling the vehicle-pedestrian contact used a force/penetration characteristic with the contact characteristics defined within the finite element model as in this instance the only noticeable deformation was incurred by the vehicle (namely, a head-strike on the plastic cowl below windscreen). One of the limitations of the MADYMO software (especially with earlier versions, such as
would have been better to have used a combined characteristic but this type of contact characteristic was only available for contacts between finite-element models and a stress versus penetration model. Alternatively, if a multibody vehicle model had been selected a mid-point contact and user-characteristic could have been defined, but as noted by Huang et al (1994b) such characteristics can be time-consuming to determine.
Throw distance was measured using the displacement measurement function of the model, taken from the model’s sternum. Although the sternum is not at the Centre of Mass of the model, it is sufficiently close to be a convenient reference point.
The simulation matrix included the following variables:
• Pedestrian orientation varied at 45 degree increments about the vertical axis with respect to the vehicle (Refer Figure 3.8).
• Vehicle impact speed between 20 and 45 km/h at 5 km/h increments.
• Vehicle deceleration at impact taken to be either 0, 4 or 8.5 ms-2, with the latter value indicating maximum achievable braking for the vehicle on that road surface.
The simulation outputs were analysed to identify test conditions that resulted in pedestrian HIC in the appropriate range and a head strike in the correct region of the vehicle.
3.8.3 Simulation Results
The throw distance results obtained and a comparison to the prediction afforded by several traditional vehicle-pedestrian accident reconstruction equations as well as the Projectile and Sliding Equation derived in Chapter 2 are shown in Figure 3.9.
The best agreement is seen for the data resulting from scenarios modelling the vehicle decelerating at -8.5ms-2, which would appear to indicate that Searle’s Equation assumes heavy braking (not an unreasonable assumption in the majority of cases).
Vehicle Impact Speed versus Throw Distance for a Typical Vehicle
0
0.00 10.00 20.00 30.00 40.00
Throw Distance (m)
Impact Speed (m/s)
Side-on to vehicle, 0 m/s^2
Side-on to vehicle, -4 m/s^2
Side-on to vehicle, -8.5 m/s^2
Facing away, 0 m/s^2
Facing away, -4 m/s^2
Facing away, -8.5 m/s^2
Facing away at 45 degrees, 0 m/s^2
Facing away at 45 degrees, -4 m/s^2
Facing away at 45 degrees, -8.5 m/s^2
Searle 93
Collins
Projectile and Sliding
Wood's Fwd Proj - Low
Wood's Fwd Proj - High
Figure 3.9 Throw Distance Comparison between MADYMO, Several Traditional Equations and the Projectile and Sliding Equation Derived in Chapter 2 versus All Results
The long throw distances apparent for the data resulting from scenarios modelling the vehicle travelling at constant speed resulted from the pedestrian being carried some distance by the vehicle before falling off, particularly at lower vehicle speeds. With the pedestrian facing away from the vehicle the collision at 6.94 ms-1 resulted in a longer throw distance than the impacts at 8.3 and 9.72 ms-1.
In instances where the vehicle was braking moderately, reasonable agreement between the MADYMO results and Collins’ and Wood’s Forward Projection (Low Estimate) equations is seen.
Figure 3.10 shows the proportion of total pedestrian travel post-impact that is airborne following an impact with a typical vehicle, versus the three different pedestrian orientations analysed. In comparison to the results from the previous section, where the collision was analysed using only a single pedestrian orientation the scatter in this instance is considerable.
The results where the vehicle was braking moderately show the greatest spread, ranging from an airborne travel proportion of 0.43 to 1.0. Where the vehicle was
travelling at constant speed the range is the narrowest, covering 0.7 to 1.0. For a heavily braking vehicle the range was from 0.57 to 1.0.
In regard to pedestrian orientation the ‘facing away’ and ‘facing away at 45º’
orientations typically produced the lowest proportions of airborne travel, particularly when the vehicle was braking moderately. The ‘side-on’ orientations produced the highest proportions of airborne travel, particularly when the vehicle was braking moderately which resulted in 100% airborne travel regardless of vehicle speed. When the vehicle was braking heavily the ‘side-on’ orientation produced a range of airborne travel proportion of between 0.8 and 1.0 whilst for constant vehicle speed the range was from 0.75 to 1.0. The average airborne travel proportion was fairly consistent, ranging between 0.82 and 0.88. This is very similar to the average airborne travel proportion determined in the previous section for a decelerating SUV-type vehicle (0.83 to 0.89).
Airborne Travel Proportion of Total Throw Distance
0.00
Figure 3.10 Pedestrian Airborne Travel Proportion of Total Throw Distance for a Typical Vehicle versus Three Different Pedestrian Orientations at Impact
It is apparent from the kinematics resulting from the simulations that for the side-on orientation the pedestrian, as ‘it’ wrapped around the front of the vehicle, travelled
away at 45º’ orientations. As noted by Simms and Wood (2005) this appears to relate to a higher effective radius of rotation about the leading bonnet edge for the ‘side-on’
orientation, leading to a larger wrap-around distance. In these instances, where the pedestrian travelled further along the bonnet than for the other orientations, it would be expected that the pedestrian would take longer to contact the ground and although it is referred to here as ‘airborne’ travel, a good proportion would actually be ‘bonnet carry’.
Generally, pedestrian orientation can be seen to influence to a considerable degree the motion resulting from a vehicle pedestrian collision.
Figure 3.11 shows the launch velocity for the pedestrian as a proportion of initial vehicle speed.
Airborne Pedestrian Velocity as a Proportion of Initial Vehicle Impact Speed
0.00 1.00 2.00
-9 -6 -3 0
Vehicle acceleration (m/s2)
Proportion
Facing away
Side-on to vehicle
Facing away at 45 degrees
Average
Figure 3.11 Airborne Pedestrian Velocity as a Proportion of Initial Vehicle Impact Speed for a Typical Vehicle versus Three Different Pedestrian Orientations at Impact
The velocity imparted to the pedestrian appears to increase with decreasing vehicle braking indicative of a shorter duration of contact and reduced energy transfer for collisions involving a heavily braking vehicle. For the averaged results, the pedestrian velocity as a proportion of vehicle velocity ranged between 0.69 and 0.82 for a heavily braking vehicle (more than -4.0 ms-2 deceleration). In comparison the same range for an SUV-type vehicle was between 0.53 to 0.8. It is thought that the major
cause of this difference would be the influence of the ‘facing away’ and ‘facing away at 45º’ orientations causing the pedestrian to obtain a greater proportion of the vehicle velocity due to greater conformity of the pedestrian’s body to the vehicle shape resulting in extended vehicle contact duration and hence greater energy transfer.
The ‘side-on’ orientation produced the lowest proportion of vehicle velocity imparted to the pedestrian whilst the ‘facing away’ orientation generally had the greatest. The
‘Facing away at 45°’ orientation generally fell in between these results except for the instance of a vehicle travelling at constant speed, where it not only indicated the greatest proportion but also exceeded 1.0 (i.e. pedestrian velocity greater than vehicle impact velocity). Although this particular result was not as high as that noted for the SUV-type vehicle-pedestrian collision analysis (1.01 in this instance, versus approximately 1.1 for the SUV-type vehicle) it can be surmised that a similar effect occurred as discussed in Section 3.7.
3.9 Discussion of the Results Obtained for an SUV-Type Vehicle and Those