More than four decades have passed since the introduction of safety standards for impact attenuation surfaces (IAS) used in playgrounds. Falls in children's playground are a major source of injuries and IAS is one of the best methods of preventing severe head injuries. However, the ability of IAS in prevention of other types of injuries, such as upper limb fractures, is unclear. Accordingly, in this paper, ten synthetic playground surfaces were tested to examine their performance beyond the collected head injury criterion (HIC) and maximum G-force (Gmax) outputs recommended by ASTM F1292. The aim of this work was to investigate any limitations with current safety criteria and proposing additional criteria to filter hazardous IAS that technically comply with the current 1000 HIC and 200 Gmax thresholds. The proposed new criterion is called the impulse force criterion (If). If combines two important injury predictor characteristics, namely: HIC duration that is time duration of the most severe impact; and the change in momentum that addresses the IAS properties associated with bounce. Additionally, the maximum jerk (Jmax), the bounce, and the IAS absorbed work are presented. HIC, Gmax, If, and Jmax followed similar trends regarding material thickness and drop height. Moreover, the bounce and work done by the IAS on the falling missile at increasing drop heights was similar for all surfaces apart from one viscoelastic foam sample. The results presented in this paper demonstrate the limitations of current safety criteria and should, therefore, assist future research to reduce long-bone injuries in playgrounds.
Play is vital in assisting children to develop fine and gross motor skills and to stay active and playground equipment is among the best tools for the physical and social development of children [1,2].
Playground safety standards place height restrictions on playground equipment based on the impact test performance of the surfacing at playgrounds. There are two pass/fail criteria for playground surfaces called head injury criterion (HIC) and maximum acceleration (Gmax). Gmax and HIC are common in all playground safety standards (except the European Union's EN 1177 IAS Standard that currently uses only the HIC ). All of the standards set the threshold of Gmax equal to 200 G and HIC equal to 1000 as the upper limits for the severity of brain injury unlikely to have disabling or fatal consequences. Most of the injuries in children's playgrounds are caused by falls [4–7] with the most severe injuries to the child's head and spine. Therefore, impact attenuating surfacing (IAS) is typically installed beneath playground equipment, with the intention of limiting the incidence and severity of these injuries. In other words, compliance with the given safety standards for IAS would prevent fatalities and reduce life-threatening injuries caused by falls onto compliant playground surfacing [3,8–18].
It seems that playground standards could successfully prevent serious injuries; however, there is an increased concern about high injury rates associated with long-bone fractures and upper body injuries. A number of epidemiological studies have identified sand surface as a safer IAS as it may significantly reduce both the incidence and severity (i.e., irreversible brain damage) of playground fall related injuries [5,19,20]. Unlike loose-fill materials such as bark, the effect of synthetic surfaces in the reduction of fall-related injuries is not consistent. For instance, Mott et al. showed a statistically significant reduction in injury incidence for falls onto rubber surfaces compared to concrete and bark surfaces . However, other researchers have shown that the incidence of fractures, typically upper limb fractures, in falls onto rubber surfaces is not statistically different to falls onto dirt and grass [19,21,22].
The ability of rubber IAS in preventing long-bone fractures remains an issue of debate among researchers [1,19,21,23–26]. Mitchell et al. suggested that as most injuries are to the limb and upper body, the standards for playground equipment and surfacing should be revisited to minimize the risk of long-bone fractures . Sherker and Ozanne-Smith have supported this by suggesting that playground fall-related arm fractures require specific countermeasures for prevention, distinct from head injury prevention guidelines .
Accordingly, to assess whether there is any issue with the current playground IAS safety criteria, ten synthetic playground surfaces were tested based on ASTM F1292, beyond the typically applied HIC and Gmax thresholds. As a conclusion, additional safety criteria are proposed which can potentially filter hazardous rubberized playground surfaces that currently comply with the HIC 1000 and 200 Gmax injury thresholds. Further research is required to set a safety threshold limit for the proposed additional safety criteria.
All the equipment used in this experiment was previously developed and verified for compliance to ASTM F1292. All tests assumed ideal conditions with no consideration for incorrect installation, aging, or degradation of the product.
The test missile had a mass of 4.6 kg and 160 mm diameter hemispherical shape. Three identical accelerometers were mounted at the center of mass of the missile. They were tri-axially arranged, so that when the missile was suspended from the release mechanism with the missile hemisphere orientated to the ground, the X and Y channels were in the horizontal plane, and the Z channel was in the vertical axis. The accelerometers were Endevco type 7264B-500 and are piezo-resistive, chosen for their DC response characteristics and simplicity of interfacing to the data acquisition system (Fig. 1(a)). The sampling rate was 25 kHz for each channel and a vector sum was performed to calculate the total acceleration. The accelerometers were calibrated against a reference accelerometer before and after the testing. A typical impact plot of time versus acceleration is shown in Fig. 1(b). The initial and exit velocities are determined by integrating the impact acceleration. An analog anti-aliasing low pass filter was used in the hardware and a digital low pass filter in the software. The data acquisition system met the requirements of ISO 6487 channel class 1000 .
Samples with different material properties were selected to provide a broad spectrum of results (Table 1). Sample codes are used in this report to describe each material. The first part of the sample code is the product identification which includes either the product color or a unique identification marking on the sample. The final number in the sample code is the thickness of the sample in millimeters. For example, Green 03-130 is a green colored 70% rubber and 30% polyurethane foam mix IAS labeled 03 with a thickness of 130 mm.
The testing procedure involved starting drops from a height of 0.5 m, then increasing by 0.5 m increments up to 3.0 m. Ten drops were performed at each height and the average is reported herein. The V8-100 sample was a very soft open-celled foam. No results were recorded for the V8 sample at 0.5 m drop height due to the low Gmax value that led to unreliable triggering of the data acquisition equipment. Additionally, drops for the ethylene propylene diene monomer (EPDM-100) were not performed at the 3.0 m height as the Gmax values were very high and may have resulted in damage to the measurement equipment.
Results and Discussion
In this section, results of the analyzed acceleration data are reported. HIC, Gmax, maximum jerk (Jmax), and maximum bounce were calculated by the data acquisition software. The work done by the IAS and the impulse force were postprocessed using matlab R15.
Current Safety Criteria, Head Injury Criterion and Gmax.
where a is the total acceleration and t1 and t2 are two intermediate values of t.
The values of HIC and Gmax obtained for each sample at various drop heights are shown in Figs. 2(a) and 2(b), respectively. At each drop height, the EPDM sample produced the highest HIC while V8-100 showed the lowest HIC values. HIC appeared to be influenced by the thickness of the rubber samples (Fig. 2(a)). The rubber IAS samples PI-50, PI-75, Blue-70, Green-70, EPDM-100 all exceeded the 200 G pass/fail threshold occurring beyond the 1000 HIC fail threshold (Fig. 2(b)).
Impulse Force (If).
where vb represents the bounce velocity (it is equal to the final velocity of the object during the impact when the object bounces back up) and vi represents the impact velocity (it is the initial velocity during the impact when the object impacts the underneath surface). In the present work, HIC Δt is derived from the calculation of the maximum HIC and is the time between t1 and t2. The shorter the HIC duration the more severe the impact and this is the rationale for using the HIC duration as the time interval. It can be seen from Fig. 2(f) that the thicker the samples the lower the impulse force. For example, PI-50 has a significantly higher impulse force than PI-75 and PI-100. PI-50 also has the worst performance of all the samples, even including EPDM-100 that was chosen as the boundary performance test sample in this study.
where J is the jerk, a is acceleration, and t is 4 μs which is determined from the sampling rate of 25 kHz. To reduce the noise within the jerk data, the values were smoothed using a 25-point running average, which corresponds to a 1 ms averaging period. The 1 ms averaging period had been previously ascertained by way of a sensitivity analysis and was chosen as the period that smoothed the data without significantly altering the value of Jmax. The maximum value for jerk for each sample at various drop heights is shown in Fig. 2(c).
The relative difference between Jmax for each sample is greater than for the HIC or Gmax for a given height (Figs. 2(a) and 2(b) versus Fig. 2(c)). For example, from a drop height of 2.0 m, the PI-50 sample has a jerk of 75,000 g/s, that is, 258% more than the PI-75 sample and 516% more than the PI-100 sample. Alternatively, for the same drop height of 2.0 m the PI-50 sample has a HIC of 1867 that is 174% more than the PI-75 and 316% more than the PI-100 sample. The Regupol samples were chosen for this analysis because they are made from the same material with only variations in thickness. This trend becomes more evident as the drop height increases or the samples get thinner. Thus, as the IAS thickness is changed there will be a greater effect on the value of Jmax and HIC.
From Fig. 2(d), for a given drop height, the bounce was very similar for nine samples (despite these samples being vastly different). The exception was the V8-100 sample. It was observed during testing that the V8-100 sample had a far slower dimensional recovery from impact then all the other IAS samples. These results indicate the difference in material properties, i.e., greater viscoelasticity and the associated delayed deflection recovery of the V8-100 as compared to other samples.
where W is the work, m is the missile mass, vi is the impact velocity of the missile, and vb is the bounce velocity of the missile. The V8-100 absorbs more energy than any of the other samples. Additionally, the values obtained indicate that the rubber samples absorb the same amount of energy for a given drop height, independent of the thickness of the sample. EPDM, despite being a considerably firmer layer than the rubber samples, absorbed a comparative amount of energy.
Magnitude of Response for Each Surface at Critical Fall Height.
1000 HIC and 200 G are the pass/fail threshold limits within ASTM F1292. All the samples in this study reached the 1000 HIC limit before reaching the 200 G. Hence, using a 1000 HIC value, the corresponding value of Gmax, If, Jmax, bounce and work were calculated using a linear approximation as shown in Table 2.
From Table 2, there are unusual behaviors in the current criteria. For the same HIC, the values of Gmax, Jmax, bounce, and work are vastly different for different samples. This highlights some of the problems with the current criteria. This is particularly evident with comparisons between PI-50, PI-75, and PI-100 samples. When the PI-50 sample has a HIC of 1000, it has a Gmax of 174 G and Jmax of 47,083 G/s. However, when the PI-75, has a HIC of 1000, it has a Gmax of 149 G and Jmax of 29,175 G/s. Additionally, similar comparisons can be made for the PI-100. When PI-100 HIC is equal to 1000 it has a Gmax of 133 G and Jmax of 19,575 G/s. This is significantly lower than the values found for the PI-50 sample, despite both PI-50 and PI-100 samples having the same 1000 HIC value. Moreover, if the samples were purely evaluated on the Gmax performance criterion, then the 174 G PI-50 impact would be considered much worse than the 133 G PI-100 impact. However, if the HIC performance criterion is used then both impacts are equally dangerous because they both have a HIC of 1000. This highlights a concern with the current standards that only use the HIC and Gmax as the performance criteria.
Additional Injury Criteria.
It can be seen from Fig. 2 that the HIC, Gmax, Jmax, and If are all influenced by the samples thickness whereas the bounce and absorbed work are not affected by the thickness of the samples. Bounce and absorbed work seem to be affected more by the sample properties rather than their thickness.
All the recorded test values for the EPDM were the highest among the samples with one exception of the bounce that all the tested samples (except V8-100) showed similar values.
All the recorded test values for V8-100 were the lowest among the samples with one exception of the absorbed work. V8-100 sample showed the highest absorbed work and therefore slower recovery from the impact compared to all other samples.
The results show that HIC, Gmax, and Jmax followed the same pattern regarding the type of surface, thickness of surface and drop height, whereby at a given drop height, the thinner samples generally produced higher HIC, Gmax, and Jmax. Maximum jerk was more sensitive to material thickness in that the magnitude rise in Jmax was considerably greater than the change in HIC or Gmax for the same change in surface material thickness.
Davidson et al. showed that the amount of energy dissipated away from or returned to a child impacting onto a surface will influence the risk of sustaining a fracture . This underlies the fact that bone is a viscoelastic material meaning that the impact response is dependent on the loading rate. At high loading rates, long bone is stiffer and can tolerate a higher load before failure than compared to a lower rate of loading . The previously mentioned variables distinguished between the various IAS samples and thicknesses tested in this study. Many of the rubberized samples exhibited almost similar bounce and work done at a given drop height despite differences in material type and thickness (Figs. 2(d) and 2(e)). One exception was the V8-100 material that resulted in a lower bounce and more work done on the test missile, which closely resembles the drop test performance of loose-fill materials. Therefore, the behavior of rubber materials in achieving a higher critical fall height (CFH) by increasing thickness and extending the impact duration and displacement (i.e., lowering the loading rate), rather than absorbing more of the impact energy, may not be effective in reducing the risk of bone fractures. Accordingly, it seems that increasing the sample thickness is not always a practical safety intervention.
In other words, our findings indicate that tested rubberized samples absorb a similar amount of energy despite differences in their material type and thickness. Due to the higher critical fall height of the thicker rubber samples, these surfaces absorbed more energy, i.e., work done than corresponding thinner samples when HIC is 1000. However, the residual kinetic energy returned to the missile, which is reflected in the bounce and bounce kinetic energy, was also larger in the thicker samples, except for V8-100.
HIC Δt is the most severe period of the impact. However current criteria do not take its significance into account. Including the momentum also addresses the IAS properties associated with bounce as any bounce is added to the total missile momentum. Therefore, the advantage of using If as an additional injury criterion is that it combines these two important injury predictor characteristics that are currently not used, viz. the time over which the HIC is calculated and the missile momentum that includes the bounce velocity. Accordingly, the high incidence of other injury types, particularly upper extremity fractures, may be related to other characteristics of the impact event, such as maximum jerk, impulse force, and the absorbed work.
Injury prevention performances of ten IAS samples were presented in this study. These showed that changes in the material type and thickness influenced the HIC, Gmax, and Jmax at increasing fall heights. Although several rubber products were able to achieve a greater critical fall height by increasing thickness, the amount of energy absorbed by the samples was similar at a given fall height. Therefore, it seems that the lack of energy absorbed by rubberized IAS may be a deficiency for fall injury risk reduction of long-bone fractures and should be considered in policy decisions regarding playground surfacing. Setting specific limits for Jmax and subtraction of the bounce from the calculated critical fall height are promising suggestions for playground surfacing standard improvements and should be considered for future studies. It would also be a method that would allow the existing 1000 HIC threshold to be retained while still acknowledging the increased injury risk presented by an IAS which returns a large amount of energy to the impacted child. Moreover, considering impulse force as an additional injury criterion is beneficial as it combines two important injury predictor characteristics that are currently not used, namely, the time over which the HIC is calculated and the missile change of momentum.
The authors would like to thank Dr. Terry Brown and Mr. Chris Chapman for their support throughout this project.