Documentation:FIB book/The Effect of Helmet Design and Regulation on Head Injuries in Snow Sports

Overview

Overview

Snowsports, such as skiing and snowboarding, are a widespread leisure activity, particularly in Canada. Unfortunately, head injuries are all too common in these sports, even in recreational cases, accounting for 19% of all reported injuries, with traumatic brain injuries (TBI) being the leading cause of fatalities in snowsports. Helmets have long been used as the standard tool for preventing or lessening the severity of head injuries in these sports, however the effectiveness of helmets is still not completely agreed upon by experts.

There are a couple of different methods of analyzing head impacts in regard to Snowsports; the most common being experimental tests, such as anthropomorphic test devices (ATDs), and numerical solutions.

Another point of contention is the testing and regulation of helmets. Studies have found that the accelerations that are tested for when following certain standards, like ASTM F2040, are not low enough to accurately represent the force needed to actually cause a head/brain injury. Conversely, the speed tested for is not high enough to fully represent the conditions in which head injuries typically occur.

Representative Work

Mechanisms of Injury

In order to properly understand the effectiveness of helmets in snow sports, it's important to first determine the scenarios of injury that helmets are supposedly designed against.

Bailley et al. conducted an investigation among skiers (n = 295) and snowboarders (n = 71) who were diagnosed with TBI from an injury whilst participating in the sport to determine the common mechanisms of injury. In the study, patients were shown sketches describing the crash scenarios and impact impact locations and were asked to choose what applied to them. The most common mechanisms were from falls (54%), then collisions between users (18%), and finally jumps (15%). With falling being the most frequent mode of injury, patients were also asked to provide the kind of fall they endured. For skiers, these common falls were falling head first, falling sideways from catching an edge (ski or board causing a "tripping" motion from momentarily digging into the snow on the hill), crossing skis, and falling backwards due to imbalance. For snowboarders, it was either by falling forwards or backwards, whether it be from catching an edge or slipping on the hill. It is important to note that not only raw statistics on frequency were taken but also other metrics such as age, sex, and importantly, TBI severity. For example, whilst falls were the most frequent mechanism of injury, collisions with obstacles (13%) caused the most serious forms of TBI. Finally, it is of note that while skiers can fall in any direction (although falling head first is the most common), snowboarder falling direction that is associated with TBI is from forward or rear head impacts which is due to the mechanisms of snowboarding itself.

Nakaguchi et al. conducted a study that relates specifically to snowboarders (n = 38) on similar issues. The study took patients that suffered from snowboarding-related major head injuries and were admitted in hospitals. They found that of all snowboarder head injuries, 68% of them were due to backward falls. Nakaguchi et al. defines this injury mechanism as the "opposite-edge phenomenon" which occurs when the toe edge (backside of snowboard) catches an edge of snow and causes the user to fall down the slope and injuring the occiput. Something they note is that this mechanism is more likely to happen on gentler slopes over steeper slopes. This is because the toe edge of the board is more likely to come into contact with the slope at shallower slope angles. The reason why this injury mechanism is so prone to causing brain injury is because the opposite-edge phenomenon causes the users body to violently deviate. The large angular acceleration at the toe edge of the board causes the impact on the head to be devastating.

Data Analytics and Surveying

Numerous sources reported on the efficacy of helmets in preventing injuries, albeit with different metrics, sample sizes, and scopes. For example, one study only used admissions from a trauma center, while another used all injury reports from ski resorts . As a result, a definite conclusion cannot be made on the effectiveness of helmets in reducing injuries from these studies. However, these sources generally agree that helmets are effective at reducing minor and moderate injuries. They also agree that current testing methods are not adequate to replicate real life scenarios.

ATD Testing

Helmet testing for snow sports has always involved the use of ATDs for physical testing. ATD testing comes in many forms as there are many ways where a user can be injured while snowboarding/skiing.

A study run by Richards et al. made use of a whole body surrogate to run physical testing on the "opposite-edge phenomenon" that was described above. The testing apparatus consisted of an indoor cable system leading to a ramp of snow for the impact to happen. The ATD used was a Hybrid III ATD with a pedestrian pelvis. The ATD was fitted with all the gear a snowboarder would wear such as a jacket, snow pants, boots, and a snowboard. The set up was such that the ATD would slide down the cables leading up to the ramp. The height and speed would be set such that it reaches a speed of 8.47 m/s (typical intermediate snowboarder speed) before hitting the set up mound of snow on the ramp, causing the opposite edge phenomenon. Tests were run both with and without a helmet (to obtain comparative data.

To obtain data, panning cameras, high-speed cameras (500 frames per second), and an electronic flash trigger at the snow mound (to track when the edge of the snowboard made contact) were used. Tracking targets were also placed on the head, shoulders, and hip to measure movement in order to make kinematic analysis. Frame-by-frame analysis of the high-speed footage was used to track the 2D position of the head over time. The group found this method of testing repeatable and was able to get conclusive results from the testing.

They found that the motion of the opposite-edge phenomenon caused the impact velocity of the head to be much larger than the starting speed of the ATD. In some cases, could reach up to 178% faster than the starting speed. The average speed of head impact was 9.08 m/s normal-to-slope and 10.61 m/s resultant. Both of which are far exceeding the the 6.2 m/s evaluation criteria outlined in ASTM F2040.

The study recognize that current snow sport helmet testing does not simulate this kind of falling action, nor do they account for head rotation. Because of this, the forces involved in real accidents might be underestimated which is affecting the standards that are required for snow sport helmets.

To go into more detail on ASTM F2040, it is the standard for "Helmets Used for Recreational Snow Sports" which is widely used in America for verifying snow sport helmets. A similar standard would be the EN 1077 which is primarily used in Europe. The ASTM F2040 standard not only gives impact test requirements (most importantly the impact velocity criteria and peak acceleration requirements) for helmets but also provides information on the headform shapes that are allowed to be used when testing the helmets. It also comes complete with the "impact test line" which designates where the impacts should be occurring when being impact tested. Importantly, the standard only lists the use of a rigid anvil when obtaining results. The main information to be pulled from this standard is as follows:

Impact Velocities

• Flat anvil - 6.2 m/s
• Hemispherical anvil - 4.8 m/s
• Edge anvil - 4.5 m/s

Peak Acceleration Requirements

• Peak acceleration must not exceed 300g

Note that in the EN 1077 standard, the peak acceleration requirement is similar at 250g.

Computer Simulations

The use of computer simulations in injury biomechanics proves useful as it allows for cheap and fast acquisition of data. However, simulations do require physical validations in order to be properly trusted. From the research we have gathered, computer simulations mainly make use of rigid body kinematics with multibody models.

Bailley et al. uses a rigid body approach to simulate snowboard impact conditions in the specific case where the snowboarder falls backwards. (opposite-edge phenomenon). They started with a multi-body model consisting of rigid bodies that were connected with mathematical joints. This model has previously been used and verified in the past for pedestrian, cyclist, and motorcyclist accident reconstructions. To use this model, they modified it to include the snowboard, boots, and bindings (mechanism that attaches boots to the snowboard). The size and standard mass of these components were also added to make the simulation more accurate. They also added a limit and constraint on the range of motion for the ankles since snowboarders have their feet locked to each other by a set distance and angle due to being attached to the snowboard. Note that the model does not include a helmet. Finally, properties of the individual bodies and joints were defined based on biomechanical data in order to give the simulation biofidelic body kinetics.

In order to characterize the snow in the model to make a distinction between "soft" and "hard" snow, they conducted experimental drop tests by using a 6kg headform with triaxial accelerometers. The two stiffness types of snow were defined as follows:

• Soft Snow - hand hardness index 2, was freshly groomed
• Hard Snow - hand hardness index 5, was groomed then frozen overnight

In order to characterize the different impact speeds, they ran trials by dropping the heads from different heights. Within the numerical model, they reconstructed the physical test by dropped the head from the same heights on the snow and validating the virtual results with the physical ones.

The model takes a set of input accident parameters such as morphology, tangential speed, body position, and snow stiffness and output final values such as the HIC and Combined Probability, CP, (obtained from the linear and rotational accelerations). For validation, the study compares the results against the Richards et al. paper mentioned in ATD Testing above. Since both are results for the case where the snowboarder falls backwards and that the original model is widely used in pedestrian injury simulation, the simulation has be adequately validated.

With 324 simulations found that the 80g was the average peak acceleration of the head and that the mean impact speed across all the sims was 7.8 m/s. Among these sims, analyses of the HIC and CP showed that 70% had a high risk of mTBI and 15% had a high risk of severe TBI. This is significant since the values for acceleration and impact speed are quite off from threshold requirements in the ASTM F2040 standard.

A separate paper by Bailley et al. also uses multibody models but this time investigates crash scenarios for both skiers and snowboarders. The model also uses the same starting human-body model but now incorporates new changes to account for the body of a skier. Angle constraints on the ankles of the skier were also given with a limit from -4° to 20° in flexion/extension. Skis also have a binding release system that snowboarders do not have which was implemented into the model based on the skier's weight and the recommended release setting for intermediate skiers. The poles of the skier were also included with an arbitrary release requirement of 50 N·m which is only there to account for the skier reflexively releasing their poles when falling. Otherwise, similar properties were added to ensure the biofidelity of the model.

The new impact scenarios included in this study are as follows:

• Forward skiing
• Sideward skiing
• Backward snowboarding
• Collisions between users (skiers only)
• Collisions with obstacles (skiers only)

They determined that collisions with obstacles are typically caused by a fall prior to the collision. This helped standardize the initial conditions as they simply added a large nondeformable object after the fall from the skier.

Among the 1149 simulations that resulted in a head impact, they came up with three significant points of contention for the impacts.

1. Forward and sideward skiing resulted in smallest impact velocities and peak linear accelerations.
2. Backward snowboarding and collisions between users always had high impact velocities (7.22 m/s) and linear head accelerations (80g).
3. Collisions with obstacles resulted in the highest head accelerations (626g).

Note again, that in point 2 the head accelerations are once again below the threshold for the pass/fail criteria in ASTM F2040 but above the published threshold for mTBI.

Bailley et al. conclude that the results of their testing shows that the impact conditions for helmet standards need to be modified as currently, they do not specify a degree of protection enough to effectively reduce the risk of TBI.

Rotation-Dampening Systems

When it comes to the use of dedicated injury prevention technologies in snow sport helmets, rotation-dampening systems such as Multidirectional Impact Protection System (MIPS) and WaveCel have not yet been integrated.

This is a topic that DiGiacomo et al. were interested in studying and compared the impact performance of a standard snow sport helmet against helmets that had MIPS or WaveCel.

For context, MIPS works to reduce rotational forces from being transferred onto the users head during an impact. This is useful in cases where an impact would be sustained at an oblique angle and reduces rotational accelerations to the brain. It works by integrating a small liner (coined the "MIPS liner" between the hard shell and the comfort padding that contacts the users head. The MIPS liner allows the helmet lightly slip when experiencing rotational acceleration, effectively acting as a dampener between the helmet and the head.

WaveCel operates on a similar principle by installing a much thicker liner within the shell of the helmet. Due to its thickness and shape, it is able to reduce both linear and rotational accelerations being sustained to the users head upon impact. Depending on the mix of forces being sustained, the WaveCel liner will flex, crumple, and shear to dampen the resultant forces reaching the head.

DiGiacomo et al. use a standard set up for testing impacts on the Hybrid III head. By dropping the neck and head surrogate on a 45° rigid plate, they collected acceleration data on all three helmets (standard snow sport, MIPS, and WaveCel). Three impact locations were used which were for a front, side, and rear impact on the headform. Note that testing was all done according to ASTM F2040 standard. This includes the impact speeds of 6.2 m/s and 4.8 m/s for flat and hemispherical anvils respectively. Since, this experiment heavily relied on testing rotational accelerations, a nylon stocking was fitted over the headforms to reduce the surface friction (Hybrid III surface friction coefficient is over two times the coefficient on a human head).

The results of the experiment were strongly in favor of implementing rotation-dampening systems in snow sport helmets. When comparing acceleration to the standard helmet, both MIPS and WaveCel technology significantly reduce what is experienced during the impact with WaveCel decreasing it by 50% for rear impacts at 6.2 m/s. For linear acceleration, understandably MIPS shows little effect but the WaveCel helmet significantly decreases the head acceleration for side and rear impacts.

WaveCel reduced the estimated probability of concussion to ~1%-2% across all impacts (front, side, and rear). This is significant since the probability is ~9.5%, ~28%, and ~84% (probabilities are read off of a chart) for front, side, and rear impact respectively at the impact speed of 6.2 m/s.

This study shows that the implementation of rotation-dampening systems in snow sport helmets should be considered in testing standards to significantly increase their effectiveness at reducing brain injury. These systems should at least see some use in snow sport helmets since it is not common compared to their use in other recreational helmets.

Snow Impact Testing

The majority of helmet testing is done using standard procedures that involve an anvil or rigid surface for impacts. This makes sense since these tests are easily repeatable and standardizable. However, it is important to see what kind of results occur from using real snow as the impact surface to see if there are any improvements to be made to helmets.

Bailly et al. conducted experiments to determine the head accelerations during impacts with snow. They were investigating how the peak head acceleration was affected both by the type of snow was impacted (soft, hard, or very hard) and whether or not the head was helmeted or not. For the helmeted tests, a standard ski helmet was used and noted that the liner within the helmet was made of a polystyrene foam which had the purpose of deforming upon impact to absorb more energy. The experiment resulted in the following values.

Mean peak head linear acceleration:

• Soft snow - 51g (±6g)
• Hard snow - 106g (±299g)
• Very hard snow - 170g (±27g)

The impact speed for these was recorded at 6.1 m/s. Note that this is aligns with the impact speed for flat anvil helmet testing outlined in ASTM F2040.

Of the 96 impacts that were tested, 85 (89%) exceeded the head acceleration and HIC thresholds for concussion and 28 (29%) exceeded for skull fracture. When comparing results between helmeted and non-helmeted impacts, the only significant result was that the helmeted headforms penetrated through less snow than the non-helmeted headforms. There were not any significant differences in linear head acceleration. These results suggest that further testing with snow impacts should be used to highlight stark differences between testing on rigid and soft surfaces. It also suggests that todays standards for snow sport helmets are not properly specified to protect users against mTBI and TBI.

Limitations and Controversies

Across all the literature that we have reviewed, the underlying controversy is that the commonly used standard for helmet testing, ASTM F2040, is not representative of real life injury mechanisms. We previously listed some details from the standard but the main points are as follows:

• Impact Velocity for flat anvil - 6.2 m/s
• Peak acceleration must not exceed 300g

As summarized above, the analysis done by Bailly et al. produced some interesting results for the case of a snowboarder hitting their head when falling backwards. They compared the numbers they got with what is listed in ASTM F2040 which is the American industry standard for headform shapes as well as specifications for the anvil that the helmet will strike. Recalling the numbers, the mean normal head impact speed used in their simulations was around 7.8 m/s while the speeds used in the ASTM standard range from 4.5 to 6.2 m/s. The peak head acceleration was also found to be largely off from the simulation where 13.8 g (G-force) was used when the pass/fail criteria in the ASTM standard is 300 g.

Future Research

Based on the previous research, most come to the conclusion that snow sport helmet standards should be re-evaluated to capture more realistic injury scenarios. A lot of research has tested and agrees on changes being made to the impact velocity and peak head acceleration requirements but much less take into account the presence of snow or non-rigid impact surfaces for testing helmets. Bailley et al. shows an isolated experiment of drop testing headforms on snow and collected velocity and acceleration data. But the use of soft surfaces for testing should be widespread across all forms of snow sport helmet testing. While this would increase the resource cost for any general testing, Bailley et al. and Richards et al. have shown that there is significant merit and including these parameters. As an example, DiGiacomo et al. explore the effectiveness of adding rotational dampening systems to snow sport helmets and use a rigid anvil for their impact testing. Including the effect of different snow hardness could give additional insight on this argument.

Regarding the ASTM F2040 standard, the impact test line is defined as positions where the helmet should be impacted for velocity and acceleration testing. With the research conducted by Bailley et al., they show that the mean impact velocities and peak head accelerations largely vary depending on the injury mechanism. But since the injury mechanism is directly tied which side of the head experiences the impact, we believe it could be possible to explore a scale of impact speed and peak head acceleration requirements depending on where along the impact line is being struck. Even though this means helmets would have to be tested up to 4 times as much (since needing to be tested at specified locations), we believe it could contribute to increasing helmet effectiveness.

Finally, the testing of collisions with other users and obstacles are very few compared to falling injuries. This makes sense as collisions between users far less easy to standardize. However, it may be useful to begin investigating these edge cases in the pursuit of increasing helmet effectiveness.