Heading out the door? Read this article on the new Outside+ app available now on iOS devices for members! Download the app.
Helmet researchers work to improve the outcome of anybody who takes a hit to their head, regardless of race, creed, gender, sexual orientation, or political leaning. Perhaps no one is more passionate about protecting human brains than Peter Halldin.
One case in particular grabbed his attention. When a motocross crash put a rider in a coma for a year despite wearing state-of-the-art body protection and a full-face helmet, Halldin was inexorably drawn to understand why. This rider was leading a motocross race on a lumpy, serpentine course repeatedly doubling back on itself, when he was suddenly hit from the side by a flying motorcycle. The throttle on that motorcycle had gotten stuck, and its rider, who was far back in the pack, couldn’t make the next turn and zoomed crosswise between back-and-forth folds of the course. Becoming airborne as he crossed the course in front of the leading riders, his front fender hit the right side of the race leader’s helmet. Though receiving no fractures, lacerations, or abrasions, its wearer suffered massive brain injuries and a coma.
His helmet did what it was designed to do; it prevented skull fracture, thanks to the thick layer of energy-absorbing EPS foam surrounding the rider’s head. However, the hematomas (swelling of clotted blood) the rider suffered in his frontal lobe and rear part of his brain rendered those areas largely nonfunctional for a long period.
Halldin, while a postdoc assistant professor at the Royal Institute of Technology (KTH) in Stockholm, Sweden, endeavored to understand how to prevent such traumatic brain injuries (TBIs).
Using a 3D model of the human brain developed by Professor Svein Kleiven at KTH, Halldin and Kleiven employed finite element analysis (FEA) to reconstruct the strain (or stretching of brain tissue) the rider’s brain experienced during the crash. The near-perpendicular impact of the fender into the helmet caused rapid rotation of the brain, which the FEA model predicted would heavily damage two areas — the exact two areas that the CT scan of the rider’s brain showed to be most damaged.
Sagittal (head nodding) motion affects the brain differently than does rotational motion. It’s the reason woodpeckers don’t get concussions from banging their beaks repeatedly into trees. Studies of boxers have shown that knockouts almost never happen after jabs; rather, they come after uppercuts, hooks, and roundhouse punches that rotate the head. This is the recipe for brain injury; abruptly rotating the brain can literally tear its internal structure, and most head injuries in bike, motorcycle, ski/snowboard, and equestrian crashes involve an oblique blow that rotates the head.
The brain is surrounded by cerebrospinal fluid; it allows the brain to slide inside of the skull, protecting it from sudden rotation on impacts. Halldin, Kleiven, and neurosurgeon Hans von Holst realized that a system built into a helmet that reduces the coefficient of friction (CoF) between the brain and what it’s encased in, the way cerebrospinal fluid does, could minimize injuries from oblique head impacts.
They created a two-layer system in which the helmet’s inner layer stays loosely coupled to the head on impact, while the outer layer slips easily on it in any direction, allowing the helmet to rotate rapidly while the skull rotates less rapidly. MIPS (Multi-directional Impact Protection System) was born, and dozens of helmet manufacturers incorporate it into their helmets.
The blink of an eye takes 100 milliseconds, and the critical moment of impact to the helmet in a bike crash is on the order of 5-10 milliseconds. During that brief moment, MIPS allows the helmet to rotate 10-15mm relative to the skull and, in so doing, lessens brain rotation, thus reducing internal damage, bleeding, and concussion.
Though MIPS was the first in this space, it’s far from alone now.
>Rather than using the MIPS slip layer, some helmet manufacturers instead have developed other methods also intended to allow free rotation of the helmet on the head upon oblique impact. For instance, some Bontrager/Trek helmets have a collapsible cellular polymer structure called WaveCel in place of much of the EPS foam.
Folds in the WaveCel cells not only absorb axial impacts by crumpling; they also fold over when subject to shear forces from oblique impacts. POC retired its silicone gel–like slip membrane called SPIN (Shearing Pad INside); its cycling helmets as of 2022 re-incorporate MIPS. Kali’s Low Density Layer (LDL) consists of viscoelastic padding placed throughout the interior of the helmet resembling octopus suction cups intended to bend upon rotation. 6D’s Omni-Directional Suspension (ODS) suspended dual-liner assembly is designed to displace and shear omni-directionally when subjected to impact.
Helmet Testing: MIPS
Determining helmet safety requires testing, and we tested six helmets beyond the certification tests required to sell bicycle helmets.
Despite the fact that head impacts in cycling almost always occur at an angle, bike helmets are only required to pass linear-impact tests in order to sell them in most countries. CPSC (USA) and CEN (EU) tests involve dropping a helmet onto a hard, flat surface with a headform inside similar in weight and shape to a human head and wired with at least one accelerometer to measure its deceleration on impact. The crushable expanded polystyrene (EPS) liner effectively reduces the head’s linear acceleration in most helmets.
While certification requires dropping helmets onto different sections (top, sides, front, back), it doesn’t simulate a typical bike crash involving striking a hard surface at an angle.
However, angling the anvil onto which the helmet with the instrumented headform inside is dropped (and covering it with sandpaper) will cause the helmet to rotate when it hits. Halldin has been working for adoption of this test protocol, using a 45-degree angled ramp, into the required international helmet certification standards, which are revised every five to 15 years.
Our MIPS Helmet Test
The cutting-edge Virginia Tech Helmet Lab, which produces the bicycle STAR helmet rankings, tested six helmets for VeloNews using a method to assess both linear and rotational impacts on the brain. Three helmet models, each in a MIPS and non-MIPS version, were tested: the Rudy Project Racemaster, the Lazer Z1, and the Scott Spunto junior helmet. (Since the non-MIPS anti-rotation helmet technologies are integrated with the padding, we could not do a comparison test with any of them. Helmets with those systems are fundamentally so different from those without that multiple variables would change.)
Virginia Tech’s STAR bicycle-helmet impact-test methodology involves dropping a NOCSAE instrumented headform encased in a helmet onto a steel, 45-degree-angled plane covered with 80-grit sandpaper (replaced after every fourth test). This evaluates a helmet’s ability to reduce not only linear acceleration but also rotational velocity of the head resulting from a bike crash.
Virginia Tech averages together brain-injury risks from different helmet-impact test configurations to compute its Summation of Tests for the Analysis of Risk (STAR) values. The STAR equation resolves the results from a range of helmet tests into a single number (1-5 stars); it was originally developed to estimate the incidence of concussion of a college football player wearing a given helmet over the course of a season.
The original version of the STAR injury-risk equation was developed by pairing concussion diagnoses in collegiate football players with head-impact data collected from sensors inside their helmets. For STAR testing of bicycle helmets, that injury-risk equation was modified to include angular velocity, which has a strong correlation with strain development in the brain leading to concussive injury.
Results: MIPS Helmet Testing
The Virginia Tech Helmet Lab performed two STAR tests on each helmet — one where the helmet hit on the side (location 1), and the second where the helmet hit on the temple area on the opposite side (location 2).
Speed at impact was 4.8 meters per second (10.7mph). The NOCSAE headform’s three accelerometers recorded the peak linear acceleration (PLA) in g’s, and its triaxial angular rate sensor recorded the peak angular velocity (PAV) [analogous to peak rotational velocity, or PRV]. The risk of concussion was calculated by inserting the PLA and PAV values into the injury-risk equation. For example, a risk value of 0.10 corresponds to a 10% risk of concussion.
With every helmet, the PLA, PAV, and resultant concussion risk were lower with the MIPS model than with the non-MIPS model at both impact locations. The biggest difference in risk was with the Scott Spunto at side impact (location 1), where the calculated concussion risk with the MIPS helmet was less than 16% of the non-MIPS Spunto.
The smallest reduction in risk with the MIPS layer was with the Rudy Project Racemaster, also at impact location 1; the calculated concussion risk with the MIPS layer was 74% of the risk without MIPS. On the Lazer Z1, the concussion risk was cut approximately in half on side impact (location 1) with the MIPS layer and by about two thirds on temple impact (location 2).
Predicting how well the helmet will protect its human wearer requires testing that faithfully simulates what happens in real-world impacts. This requires modeling more interfaces than current helmet testing includes.
Simply angling the impact surface and making its coefficient of friction (CoF) similar to that of asphalt to simulate what happens at the helmet/road interface in a bike crash is not enough. Faithfully modeling the interface between the skin (and/or hair) and the helmet as well as the interface between the skin and the skull is also required; current headforms fail to do so.
The three headforms most commonly used for impact testing of sports helmets are the Hybrid III (aka “HIII”) headform, developed for automotive crash testing, the magnesium EN960 headform used in European motorcycle-helmet tests, and the NOCSAE (National Operating Committee on Standards for Athletic Equipment) headform, created for testing helmets for football, hockey (ice and field), baseball/softball, polo, and lacrosse.
Other than the magnesium EN960, these headforms have either a vinyl or polyurethane layer over them. That surface is first of all bonded to the headform (unlike human skin, which slides on the skull) and is secondly stickier than human scalp, with or without hair. Some studies have found the CoF of those headforms to be around four times as high as that of human scalp and hair. So, if you test a helmet incorporating a system designed to reduce the coefficient of friction between helmet and head on a headform that is stickier than a human head, will the result be representative of what happens in the real world?
MIPS Helmet Testing in the Real World
Antonia Trotta from the University College Dublin led a group publishing a paper in the June 2018 Annals of Biomedical Engineering studying exactly this. The authors quoted tests on cadaver heads (without helmets) showing that, “the impact force undergone by the head without the scalp is up to 35% higher than with the presence of scalp, depending on the impact angle between the impact surface and the head.”
They superglued pig scalp similar in thickness and properties to human scalp to the headform at the same anatomical connection points that the skin on a human skull is attached. The group dropped the uncovered and pigskin-covered headforms with and without a helmet onto both flat and angled anvils. They noticed little change in linear acceleration on the flat anvil with the simulated scalp.
However, the porcine scalp created large reductions in rotational acceleration and angular velocity of the brain on the angled anvil. Their studies also showed that the higher the friction of the angled surface the helmet hits, the greater the difference in rotational acceleration and angular velocity of the simulated brain between a headform with and without simulated scalp. The presence of hair was not considered.
Stephanie Bonin, a senior biomechanical engineer in the Injury Biomechanics Group at MEA Forensic, and two colleagues followed this study up in 2020 by covering an HIII headform in two nylon stocking layers to reduce its CoF. They dropped helmets both with and without a MIPS layer inside containing both bare and stocking-covered headforms onto a 45-degree anvil with 40-grit sandpaper glued to it.
Both the MIPS layer and the stocking layers made only tiny changes to peak linear acceleration (PLA). However, both the MIPS slip-plane layer and the stocking layers significantly reduced peak angular acceleration (PAA) and peak angular velocity (PAV).
Equally interesting: the MIPS layer made a similar reduction in these measurements, whether the stocking layers were present or not. So, the lowest PAA and PAV readings, by far, were with both the MIPS layer combined with the stocking layers. Bonin also tested human hair, published in 2021, which further reduced PAA and PAV, and, again MIPS reduced PAA and PAV yet further on the hair-covered headform.
You might wonder why the neck and body is not simulated in these helmet tests. Halldin states two main reasons why they are not included in helmet tests. The first is that the only dummy neck available is a Hybrid III neck designed for frontal automobile crash simulations, and it does not behave like a human neck in oblique impacts.
He also theorizes that during the 5-10 milliseconds of the damaging rotational acceleration of the brain, the neck (and body) have insufficient time to affect the head. Virginia Tech originally experimented with the HIII neck, as did the designers of Trek/Bontrager’s WaveCel technology; both have since transitioned to testing helmets without an attached neck.
Results on the MIPS website are all with the HIII headform, which Halldin believes is more biofidelic (faithfully models biology) than the EN960. He has measured the coefficient of friction (CoF) for four headforms. It is in the 0.6-0.7 range for both the HIII and the NOCSAE headforms. The EN960 headform has a CoF of 0.16-0.18, and the new headform from WG11 has a CoF of 0.3.
Virginia Tech’s STAR tests are performed using the NOCSAE headform.
MIPS Helmet Testing: Into the Technical Weeds
To understand the risk of brain injury in the Virginia Tech STAR ratings, it’s important to understand AIS and BrIC.
The Abbreviated Injury Scale (AIS) gives a 1-6 numerical ranking to the severity of injury to a given part of the body, with 6 being fatal (AIS 1 = Minor, AIS 2 = Moderate, AIS 3 = Serious, AIS 4 = Severe, AIS 5 = Critical, and AIS 6 = Maximal). For the brain, AIS 2 is a concussion; AIS 3 and above is TBI.
Based on finite-element modeling of human brains, Rotational Brain Injury Criterion (BrIC) corresponds to a Cumulative Strain Damage Measure (CSDM) for each AIS. The BrIC equation relates maximum and critical rotational velocities and maximum and critical rotational accelerations to come up with a number that determines brain-injury risk. Concussive (AIS 2+) BrIC values for humans varied from 0.60 from animal data to 0.68 using data from college football players, whereas the BrIC value corresponding to a 30% risk of an AIS 3+ TBI is 0.92.
According to a Development of Brain Injury Criteria (BrIC) study in the November 2013 Stapp Car Crash Journal by Takhounts, Craig, Moorhouse, and McFadden of the National Highway Traffic Safety Administration, rotational velocity and not rotational acceleration is the mechanism for brain injuries. Since angular acceleration did not correlate well to any physical parameter, it was excluded from the BrIC formulation. Virginia Tech includes rotational velocity and not rotational acceleration in its STAR tests.
Comparing headforms, a study by Bland, McNally, and Rowson presented at the 2018 International Research Council on the Biomechanics of Injury (IRCOBI) conference in Athens, Greece compared data in oblique helmet drop tests from HIII and NOCSAE headforms and also tested the validity of testing with the HIII neck. It concluded that, “the NOCSAE headform produced similar PLA but lower PRV and PRA compared to the HIII, while the incorporation of the HIII neck generally decreased PLA, PRV and PRA compared to head-form‐only tests.”
Comparing anti-rotation technologies between brands, in a January 2020 article in the Annals of Biomedical Engineering entitled “Impact Performance Comparison of Advanced Bicycle Helmets with Dedicated Rotation-Damping Systems”, authors Bottlang, Rouhier, Tsai, Gregoire, and Madey tested helmets with MIPS, SPIN, LDL, and ODS anti-rotation technologies against a control helmet with simple EPS padding. Their study validated MIPS and (POC’s) SPIN in reducing brain injuries from oblique impacts; it did not validate Kali’s LDL and 6D’s ODS.
Bottlang et al dropped helmets attached to a Hybrid III headform and Hybrid III neck onto a 45-degree angle anvil at 6.2 m/s. Their results showed that peak rotational velocity (PRV), peak rotational acceleration (PRA), and BrIC (brain injury risk) of MIPS and SPIN helmets were significantly reduced compared to control helmets.
Their results also showed that there was no significant difference in peak rotational velocity (PRV), peak rotational acceleration (PRA), and BrIC between ODS and LDL helmets compared to control helmets. Furthermore, peak linear acceleration (PLA) of helmets with an anti-rotation system was not significantly different from that of control helmets, except for LDL helmets, which exhibited a 62% higher linear acceleration than control helmets.
They concluded that “the probability of experiencing AIS 2 brain injury, predicted by BrIC, was significantly reduced by 42% (p = 0.002) and 54% (p = 0.003) for MIPS and SPIN helmets, respectively, compared to control helmets. There was no significant difference in PAIS 2 between ODS and LDL helmets compared to control helmets.”
They also concluded that “the Hybrid III neck was validated for flexion and extension but has been shown to be overly stiff in lateral bending and axial compression.” The study acknowledges that Bottlang and Madey “are co-inventors of a helmet technology, called WaveCel. They have filed patents, and have a financial interest in WaveCel,” which is licensed by Trek/Bontrager.
New Helmet Standards
The FIM (Federation Internationale de Motorcyclisme) introduced a test for Moto GP race helmets (FIM FRHPhe-01) including a rotational test where the helmet is dropped at 8m/s onto a 45-degree impact anvil. That test method has partly been copied into the new European Motorcycle standard (ECE 22.06). The FIM and ECE 22.06 test standards that include rotational testing define a Pass as < BrIC 0.68, indicating AIS < 2 (no concussion). Rotational testing for the CEN/TC 158 (European) standards for cycling, skiing, mountaineering, safety (construction helmets, etc.), and equestrian sports will follow in 2025.
The CEN/TC 158 Working Group 11 (WG11), convened by Halldin and consisting of researchers and representatives of helmet companies, test laboratories, and headform manufacturers, is developing new helmet-test standards for cycling, skiing, mountaineering, safety, and equestrian sports. The new standards will have pass/fail criteria based on injury-risk assessments like BrIC or PRV.
WG11 is in agreement on the test methodology, namely dropping a helmet-encased headform (minus neck) onto a rough-sandpaper-covered steel anvil inclined at 45 degrees, but not yet on what defines the pass/fail criteria. Within WG11, a new headform has also been developed with more humanlike inertial properties and friction properties relative to the helmet.
A criterion of the new WG11 headform is that it conforms to the coefficient of friction (CoF) of 0.3 from the 2018 Trotta et al study.
While many helmet companies aren’t waiting for new test standards that include rotational forces on the brain and are introducing models that have a slip function reducing the friction between the head and the helmet, KASK is not. KASK product manager Luca Viano says that, since no peer-reviewed test protocols exist for rotational helmet testing, there is no consistency among tests and no certainty about what constitutes a safe helmet. From rotational testing of KASK helmets using the EN960 headform (from the motorcycle standards), Viano claims that all KASK helmets test below BrIC 0.3, well below the BrIC 0.68 (AIS < 2) pass/fail limit of the new motorcycle test standards. KASK helmets do not rate high on Virginia Tech’s STAR rankings, though; Viano attributes this to the high CoF of the NOCSAE headform used for the STAR tests.
KASK has recently had oblique-impact tests done in Milan using an EN960 headform. KASK believes that the slipperier magnesium EN960 headform more aptly approximates the CoF a human head.
Due to the inadequacy of current headforms to faithfully simulate human scalp and hair, helmet testing has not evolved to the point that we can definitively say how much damage the brain will experience under a given impact in a given helmet.
That said, the existing testing clearly points to the value of wearing systems that reduce the coefficient of friction between the head and the helmet on oblique impacts.
And the testing also indicates that the greatest protection against brain damage due to oblique impacts are achieved by combining the natural slipperiness of the head’s skin and hair with a rotational friction-reducing system like MIPS.
Lennard Zinn joined VeloNews in 1987. He is also a custom frame builder (www.zinncycles.com) and purveyor of non-custom huge bikes (bikeclydesdale.com), a former U.S. national team rider, co-author of “The Haywire Heart,” and author of many bicycle books including “Zinn and the Art of Road Bike Maintenance,” “DVD, as well as “Zinn and the Art of Triathlon Bikes” and “Zinn’s Cycling Primer: Maintenance Tips and Skill Building for Cyclists.” He holds a bachelor’s in physics from Colorado College. Follow @Lennardzinn on Twitter.