Helmets and protective gear for e-scooters: crash physics, the standards matrix, rotational mitigation, and FOOSH biomechanics

19.05.2026 Scootify 15 min read

The companion guide «Safety gear, traffic rules: how to ride without ending up in hospital or paying a fine» covers the behavioral and regulatory side of safety: injury statistics, traffic codes across countries, anti-patterns, and a surface overview of helmet standards. This article is an engineering deep-dive into crash physics and the certification mechanics behind protective gear: why the Head Injury Criterion threshold is 700, not 1500; what actually differs between EN 1078 certification for a 25 km/h bicycle and NTA 8776 for a 45 km/h speed pedelec; why ASTM F1492 lets a skateboard helmet take multiple impacts while EN 1078 lets a bicycle helmet take only one; how MIPS and WaveCel physically reduce rotational acceleration of the brain, and which independent metric confirms this; why FOOSH (Fall On Outstretched Hand) accounts for 25 % of all childhood bone injuries, and why wrist guards have splints of a specific length. This is a discipline of its own, parallel to braking technique, acceleration, cornering — an applied-physics circuit of rider skills, but focused on protection rather than control.

The prerequisite is understanding how the contact patch generates forces on the wheel, what happens when contact with the surface is lost, and why emergency maneuvers and obstacle avoidance sometimes are not enough — some falls are unavoidable, and at that moment the deciding factor is no longer handling but the gear’s ability to dissipate kinetic energy over a time shorter than the biomechanical injury threshold.

1. Crash physics — linear acceleration, rotational velocity, and the injury scale

After a fall the head experiences two superimposed kinematic events that act biologically in very different ways:

Linear (translational) acceleration comes from a direct impact between helmet and a hard surface. The head — with the brain inside — decelerates from working speed to zero in milliseconds. If the peak deceleration is too high, the brain exceeds the threshold of skull compression: «focal» injury — contusion, skull fracture, bruise at the impact site.
Rotational (angular) acceleration arises when the impact is not strictly perpendicular to the surface — and it almost never is strictly perpendicular. The head’s angular velocity rises sharply, and the brain, as a viscoelastic mass, lags behind the skull’s motion, stretching axons at the boundary between grey and white matter — the mechanism of diffuse axonal injury (DAI), the molecular basis of concussion.

The standardized scale for the linear risk is the Head Injury Criterion (HIC), adopted by NHTSA into FMVSS 208 in 1972 as the integral of head center-of-gravity acceleration over a given time window:

HIC = max { [(t₂ − t₁) × ((1/(t₂−t₁)) × ∫a(t)dt)^2.5] }, where t₂ − t₁ ≤ 36 ms (HIC36) or ≤ 15 ms (HIC15)

Since 2000 the current FMVSS 208 has used HIC15 with a threshold of 700, corresponding to ≈ 5 % risk of serious head or skull injury (Wikipedia § Head injury criterion, NHTSA review). HIC15 = 1000 (the old number for HIC36 from 1972) corresponds to ≈ 50 % risk of skull fracture on the same curve. The nonlinear 2.5 exponent is a key engineering hint: biologically it is more dangerous to take one short 200 g peak over 5 ms than 100 g over 20 ms with the same average impulse, because the short peak lands in the zone where brain tissue cannot redistribute the strike fast enough.

The standardized scale for rotational risk is the Brain Injury Criterion (BrIC), adopted by NHTSA later, in the 2010s: a dimensionless sum-of-squares of peak angular velocities along three axes, normalized to critical values (≈ 66 rad/s on x, 56 rad/s on y, 42 rad/s on z for the human cradle). BrIC = 1.0 corresponds to a 50 % probability of AIS 4+ (severe brain injury). In real e-scooter crashes at 25 km/h, an oblique impact against a curb generates angular velocities of 30–60 rad/s, right in the concussion-risk zone.

A third scale is the Gadd Severity Index (GSI), a predecessor of HIC, still used in Snell. It is simpler than HIC but correlates worse with real injury; Snell sets the pass threshold at GSI 1500.

The engineering task of a helmet is to convert a sharp, short peak of impact impulse into a longer, gentler one, with the same average impulse but a lower peak. This is done by controlled foam crushing: when the foam is compressed by 50–70 % of its volume, it absorbs kinetic energy via irreversible cell collapse, and the peak deceleration drops from ~ 300–500 g (a bare skull on asphalt) to ~ 150–250 g (an EN 1078-certified helmet). This is the only physical principle underlying every current standard.

2. Helmet anatomy — three layers of protection

A certified helmet is a three-tier system, with each layer solving a different problem:

Outer shell — polycarbonate, ABS plastic, or, in the premium segment, a composite with carbon fiber. Its job is to spread a point impact over a larger area and protect against penetration by sharp objects (curbstone, spike, branch). The thin micro-shells used in bicycle helmets (in-mold) are 0.5–1 mm; motorcycle helmets carry structural shells of 3–6 mm. The shell itself absorbs very little energy — that is the job of the next layer.
Foam liner (EPS / EPP) — the main absorber. EPS (Expanded Polystyrene) is the standard material in bicycle and motorcycle helmets: 40–80 kg/m³ density, single-impact (permanently destroyed by compression). EPP (Expanded Polypropylene) is a flexible analog, multi-impact (recovers after compression), used in skateboard helmets to ASTM F1492. Composite EPS+EPP is a two-layer liner in the premium segment, where EPP handles low-energy impacts and EPS handles high-energy ones.
Retention system + comfort liner — a Y-junction chin strap plus soft pads on forehead/occiput. If the helmet comes off the head before impact, the first two layers do not matter.

BHSI’s EPS Foam Helmet Liner Performance With Age shows an interesting nuance: the EPS foam itself does not degrade significantly with age under normal storage — a ten-year-old helmet stored in a drawer absorbs impact almost as well as a new one. What degrades is the outer shell (UV), the straps (sweat and salts), and the glue between liners (heat and sunlight). The canonical recommendation to replace a helmet every 3–5 years (CPSC) or 5–10 years (Snell) is really a recommendation about shell and retention, not about the foam.

3. Standards matrix — drop heights, anvils, max acceleration

Each standard is an engineering contract: «a helmet that passes this test is guaranteed not to give the head more than X g of acceleration when dropped from height Y onto an anvil of type Z». Standards are not equivalent: they use different impact velocities and different anvil shapes. The comparison table below collects the key numbers:

Standard	Use case	Flat anvil (drop h)	Impact velocity	Curbstone / hemispheric anvil	Max peak g	Single / multi-impact
EN 1078:2012+A1 (Wikipedia § EN 1078)	bicycle, inline skates, kickboard	1.5 m	5.42 m/s (≈ 19.5 km/h)	1.06 m curb, 4.55 m/s	250 g	single
NTA 8776:2016 (XNITO)	speed pedelec up to 45 km/h	≈ 6.2 m/s (≈ 22 km/h)	150 J (vs EN 1078 ≈ 100 J)	flat + curb	250 g	single
ASTM F1492-25 (ASTM)	skateboard, trick roller	flat, cylindrical hazard, triangular hazard anvils	≈ 5.0–6.2 m/s	—	250 g	multi
CPSC 16 CFR 1203 (eCFR)	US federal bicycle	2.0 m, 6.2 m/s	1.2 m curb + hemispheric, 4.85 m/s	300 g	single
DOT FMVSS 218 (eCFR § 571.218)	US federal motorcycle	1.83 m, 5.0–5.4 m/s	hemispheric anvil	400 g	single
ECE 22.06 (RideApart)	European motorcycle	slow ≈ 6.0 m/s + high ≈ 8.2 m/s	two speeds	slow ≤ 180 g, high ≤ 275 g	single
Snell B-95 / M2020 (RevZilla)	premium bicycle / motorcycle	up to 2.0 m flat	up to 7.75 m/s	flat + hemispheric + edge	GSI ≤ 1500 (≈ 275 g HIC)	single

Key engineering takeaways from the matrix:

EN 1078 is a 25 km/h test. Impact velocity 5.42 m/s corresponds to a fall without any motor or curbstone impulse from a height of ~1.5 m. This is close to the working speed of an urban e-scooter at 25 km/h, but it does not account for the fact that the impact often happens with an additional horizontal vector. If the actual body speed is 25 km/h = 6.94 m/s and the impact vector is at 45°, then the projection on the normal is 4.9 m/s — right at the EN 1078 boundary. At 35 km/h = 9.72 m/s, EN 1078 no longer formally covers the case.
NTA 8776 is a 45 km/h test. NEN (Royal Netherlands Standardization Institute) wrote it in 2016 specifically for the Dutch speed pedelec — an electric bicycle up to 45 km/h. The 150 J impact, vs the 100 J of EN 1078, is 50 % more energy, roughly the kinetic-energy difference between 25 and 35 km/h. The 250 g cap is the same, but it has to be met at a much higher input energy. Manufacturers producing NTA 8776 helmets: Abus, Lazer, Bell, Specialized.
ASTM F1492 is multi-impact. This is the key difference from the bicycle standard: EPP foam recovers after each impact, so a skateboard helmet is designed for 30+ falls per season at a skate park, not for a single major crash. For an e-scooter this is a trade-off: city riding does produce several low-speed falls per year (curb hook-up, slip on wet), and an EPP helmet handles them; but one heavy impact at ~30 km/h — the EPP liner won’t absorb energy the way EPS would, because it gives the energy back elastically.
CPSC allows 300 g, EN 1078 only 250 g. Paradoxically, CPSC is mandatory in the US since 1999, while EN 1078 is voluntary in Europe. The engineering trade-off: CPSC has a wider coverage zone and a higher flat-anvil impact velocity (6.2 m/s vs 5.42 m/s in EN), so the absolute level of protection is roughly equivalent.
Motorcycle standards DOT / ECE / Snell allow 400 g (DOT) down to 275 g (ECE high speed). This looks worse, but the impact velocity is much higher (up to 8.2 m/s in ECE 22.06 high speed) — a motorcycle helmet is designed for falls from 60–100 km/h, where even 400 g over 6 ms is better than 800 g on a bare head.

A matrix conclusion by e-scooter working speed:

E-scooter working speed	Minimum standard	Recommended standard
Child up to 16 km/h	ASTM F1492 (multi-impact — falls are frequent)	EN 1078 / CPSC + MIPS
Urban up to 25 km/h (eKFV, ПЛЕТ)	EN 1078 / CPSC	EN 1078 / CPSC + MIPS, or NTA 8776
Elevated 25–45 km/h	NTA 8776	NTA 8776 + MIPS, or a full open-face motorcycle helmet ECE 22.06
Off-road / hyperscooter 45+ km/h	ECE 22.06 / DOT FMVSS 218	ECE 22.06 + full coverage, full-face

4. Rotational mitigation — MIPS, WaveCel, KOROYD, SPIN

None of the traditional standards (EN 1078, CPSC, DOT, ECE 22.05) measures rotational acceleration directly — they measure only linear deceleration on flat and curb anvils. This is a historical artifact: when these standards were written in the 1970s–1990s, the concept of diffuse axonal injury was not yet formalized, and biomechanical models did not allow a numeric rotational threshold. ECE 22.06 (2022) and Snell M2020 are the first standards with partial rotational tests; the rest are catching up.

In 1996, Swedish neurosurgeon Hans von Holst and biomechanics scientist Peter Halldin at Karolinska Institute developed MIPS (Multi-directional Impact Protection System), a passive rotational mitigation technology. The principle: between the EPS foam and the comfort liner sits a thin low-friction plastic layer that lets the helmet shell rotate relative to the head by 10–15 mm in any direction during an oblique impact (Wikipedia § MIPS, BHSI § MIPS). This proximal motion dissipates rotational energy through friction at the slip plane before it is transmitted to the brain. Independent testing shows a reduction in rotational acceleration of up to 50 % versus conventional helmets.

WaveCel (Trek/Bontrager, 2019) is an alternative technology: instead of an internal slip plane, it uses an engineered honeycomb structure of inverted-V cells that both collapse (absorbing linear acceleration) and shear (absorbing rotational) at the same time (Pinkbike § WaveCel). Independent CPSC-style testing has shown −16…−26 % linear acceleration vs EPS and up to 5× reduction in rotational acceleration — the highest publicly tested numbers. A WaveCel helmet still has a thin EPS shell around the honeycomb — the structure replaces part of the foam thickness, not all of it.

KOROYD is a third technology: thousands of co-polymer extruded tubes thermally welded into a monoblock. The tubular structure crumples on impact, absorbing linear acceleration with better heat transfer (better ventilation). Independent tests show KOROYD’s advantage over plain EPS is on the linear component only; it does not affect rotation, so manufacturers using KOROYD (Smith, Endura) often add a MIPS liner beneath it (Singletracks § WaveCel vs KOROYD).

SPIN (POC, 2017–2022) — silicone-pad rotational damping — was a direct competitor to MIPS. POC retired SPIN in favor of MIPS Integra after 2022 without a public explanation, but independent Virginia Tech tests showed performance close to, but not better than, MIPS. If you own an older POC with SPIN, it works, but the industry has consolidated on MIPS.

Key engineering hint: a rating that accounts for the rotational component cannot come from an EN 1078 / CPSC certificate — both allow a «pass» with 0 % rotational mitigation. To evaluate rotational protection, look at the independent 5-star STAR rating from Virginia Tech (next section).

5. Virginia Tech STAR rating — a biofidelic metric

The Virginia Tech Helmet Lab, in partnership with the Insurance Institute for Highway Safety (IIHS), developed the STAR (Summation of Tests for the Analysis of Risk) rating (Virginia Tech Bicycle Helmet Ratings, VT Helmet Lab News 2025) — a public 5-star scale evaluating bicycle helmets across 24 impact tests:

6 positions on the helmet (front, side, top, rear + rim oblique positions).
2 impact speeds (low and high).
2 anvil types (flat + 30° oblique drop), since 60–90 % of real cycling crashes involve an oblique impact.

For each of the 24 tests, both linear deceleration and angular velocity are measured. A separate formula combines them into a composite score; lower is better. 5 stars = top 50 % overall; 4 stars = top 50–75 %; below that, weaker concussion protection.

This is the only public metric that accounts for the rotational component for bicycle helmets. Manufacturers do not pay Virginia Tech (unlike Snell certification, where helmets are submitted by the manufacturer); VT buys helmets at retail. So the STAR rating is an independent third-party benchmark, not a marketing label.

In 2025 Virginia Tech raised the 5-star threshold (Bikerumor § VT STAR update) — because after a decade of public ratings, the industry had improved so much that 80 % of helmets received 5 stars, and the distinction had eroded. Now 5 stars again means «top-tier helmet by concussion biomechanics».

Practical takeaway: when choosing a helmet for an e-scooter (25 km/h and above), first verify the mandatory standard (EN 1078 / CPSC / NTA 8776), then open the VT STAR rating database and pick a 4- or 5-star one. A helmet without a VT test either has not been published yet (niche or new) or has not made the target rating.

6. Fit and retention protocol — how the helmet sits

A poorly fitted helmet absorbs energy much worse, even if it is best-in-class by standard. The canonical fit check has three steps:

Two-finger above brow rule. The front edge of the helmet should end two fingers above the eyebrow. Higher than that — forehead unprotected; lower — restricted field of view and the helmet pushes into the nose during a forward impact.
Y-junction strap geometry. The two straps of the retention system meet at a Y-junction directly under the ear (not behind the ear, not in front of the ear). This is the geometry where the helmet won’t slide off, neither backward (past the rear point) nor forward under pressure.
Two-finger under chin. The fastened chin strap leaves room for exactly two fingers flat, no more, no less. More — the helmet comes off on impact; less — it suffocates the wearer during long rides, and the user loosens it back into the previous risk zone.

Shake test: helmet on and strapped, shake the head vigorously in all directions. The helmet should not shift more than 1–2 cm.

Roll-off test (ECE 22.06) is a standardized version: try to «remove» the helmet from the rear, as if from a strike to the occiput. If the helmet comes off, the retention system failed.

Retention strap test under ECE: a 10 kg weight is attached to the fastened chin strap and dropped from 0.75 m; the attachment point on the helmet must not displace more than 25 mm (IRCOBI 1992 Helmet Retention). A separate static test — 23 kg pulls the strap for 1 minute, then 38 kg of dynamic load is added; the strap must withstand a temporary tensile strength of ≥ 3 kN.

Monthly inspection routine: strap not stretched, buckle not worn, soft pads attached to the shell, EPS liner free of visible cracks or dents (especially in the spots where the helmet often lands on the floor during storage). A helmet that has experienced an impact in a crash is unconditionally replaced — even if no defects show externally, the EPS has deformed and will no longer absorb on the second hit.

7. FOOSH biomechanics and wrist protection

The second-most frequent e-scooter injury is a broken wrist. In the Swedish fracture register 2019–2022, among 1,874 e-scooter fractures, 19 % each were hand, wrist, and proximal forearm (PMC § Epidemiology of Fractures Following Electric Scooter Injury). This is the classic FOOSH injury — Fall On Outstretched Hand: the instinctive reaction to stick out a hand before falling.

FOOSH biomechanics (NCBI § Wrist Fracture, PMC § Frykman VIII Fracture, UBC Wiki § FOOSH):

A body’s kinetic energy at 75 kg and 25 km/h = ½ × 75 × 6.94² = 1,804 J.
During an outstretched-hand fall, up to 60 % of this energy passes through the distal radius via axial impact plus bending.
The distal radius forms ≈ 80 % of the wrist joint surface, so 80 % of the absorbed impulse passes through it.
The yield limit of cortical bone in the radius is roughly 210 MPa in compression, 130 MPa in bending. A 1 cm² cross-section at the thinnest point is 13–21 kN maximum, while a real FOOSH load reaches 5–8 kN on the dry hand.

The two principal fracture models:

Colles fracture — fall with the hand in pronation (palm down). Distal fragment displaced dorsally (upward). The most common classic picture.
Smith fracture — fall with the hand in supination (palm up). Distal fragment displaced volarly (downward). Rarer, harder to treat.

Frykman classification — 8 types of distal radius injury distinguished by ulnar styloid involvement and intra-articular extension. Frykman VIII is the most severe type (extra-articular with ulnar styloid involvement) and often requires plate-and-screw surgical fixation.

The wrist guard as a physical antipattern to FOOSH:

A rigid splint (metal or composite) runs along the volar (palm) side of the forearm, from the base of the palm to 6–8 cm above the wrist.
During an impact near 45° (the typical FOOSH position) the splint restrains hyperextension of the wrist, preventing the radius from bending past the breaking point.
Energy is redirected into the forearm muscles and soft tissue, where it is absorbed at a much lower peak force.
The best wrist guards have splints on both volar and dorsal sides — to also restrain hyperflexion (Smith-type fall on the back of the hand).

The most-cited evidence of efficacy comes from snowboarders. Wrist injuries are the most frequent unit of trauma in snowboarding, and wrist guards have become almost standard gear. A meta-analysis (Springer § White Paper on Wrist Protectors) shows wrist fracture risk reduced by 50–60 % in wrist-guard wearers vs the control group.

ASTM F2040 is the standard for helmets in recreational snow sports, not wrist guards directly. The dedicated standard for wrist guards is ASTM F1849 (skating equipment); this one is poorly known and manufacturers rarely advertise it, since the market is mostly snowboard- or skateboard-oriented.

For an e-scooter rider: wrist guards in the «snowboard» or «inline skate» category (Pro-Tec, 187 Killer Pads, Triple Eight) are the most practical pick. A splint of 5–6 cm above the wrist covers the basic protection; if you want to cover the proximal radius, look for «long splint» models of 8–10 cm.

8. Knee / elbow / back armor — D3O and EN 1621

If the helmet and wrist guards are the two top-priority pieces of gear, then knee, elbow, and back protectors are the next tier — recommended for off-road and hyperscooter speed ranges, optional for the city.

D3O (Wikipedia § Dilatant, explainthatstuff.com § Energy-absorbing materials) is the commercial name of a dilatant (shear-thickening) polymer invented by British engineer Richard Palmer in 1999. The principle: at rest the material behaves like a soft gel (molecules slide freely), but under a sudden impact (high shear rate) the molecules instantly «lock» into a rigid structure, absorbing energy by distributing the force across a large area. A few seconds after impact the material relaxes back to soft, ready for the next cycle.

The physical category is non-Newtonian dilatant, described in the literature since the 1930s (cornstarch with water is a household example: you can run across the surface, but you sink if you stand still). D3O is the engineered version with controlled viscosity-vs-shear-rate parameters.

EN 1621-1:2012 is the European standard for limb protectors (shoulder, elbow, forearm, hip, knee, shin). The test: a 5 kg flat striker drops on the armor at 4.47 m/s, delivering 50 J of kinetic energy (SATRA § EN 1621, Stealth Armor § CE Level 1 vs Level 2):

Level	Max mean transmitted force	Max single peak
Level 1	≤ 18 kN	≤ 24 kN
Level 2	≤ 9 kN	≤ 12 kN

Level 2 absorbs impact twice as well at the same 50 J test. D3O and Sas-Tec are the two brands dominating Level 2 armor for motorcyclists; for an e-scooter rider at 25–40 km/h Level 1 is enough for most scenarios, Level 2 for off-road and hyperscooter.

EN 1621-2:2014 is a separate standard for back protectors (full-back, central-back, lumbar). Same test geometry, but a different striker shape and anvil — simulating impact against a hard object by ribs or spine. Same Level 1 / Level 2 scale.

How these armor categories are worn:

Hard cap (rigid plastic cap with a thick pad inside) — the Pro-Tec, Triple Eight, 187 Killer Pads category — is the most popular for skateboarding/scooter. Good for falls on concrete; sometimes the head clashes with the cap itself in a very hard fall.
Soft cap (soft armor with D3O, EVA foam, Poron XRD) — the motorcyclist underarmor category — is the most comfortable for everyday wear under clothing. Absorbs better than a hard cap at the same EN 1621-1 level.
Hybrid — a rigid cap with a D3O insert inside — is the premium segment.

Safety matrix for the e-scooter rider:

Scenario	Recommended
City, 25 km/h, smooth pavement	EN 1078 helmet + wrist guards
City, 25 km/h, cobblestone / tram tracks (see difficult road surfaces)	+ EN 1621-1 Level 1 knee + elbow
Elevated 25–45 km/h	NTA 8776 helmet + wrist + EN 1621-1 Level 1 knee + elbow + EN 1621-2 Level 1 back
Off-road / hyperscooter 45+ km/h	ECE 22.06 full open-face helmet + Level 2 knee + elbow + Level 2 back + spine protection
Courier (8+ hours in the saddle)	EN 1078 + MIPS, wrist guards, Level 1 knee + back; choose for long-wear comfort

9. Eyewear, gloves, footwear — the third tier

Eyewear. Glasses at 25+ km/h protect against insects, twigs, and small particles flying off the road. Certification: ANSI Z87.1 (US) or EN 166 (Europe) for impact-rated lenses. Ordinary sunglasses are not certified and may shatter into sharp fragments on impact. Lab-grade safety glasses pass — they are the cheapest viable option.

Gloves. Palm skin abrades through to bone instantly in a FOOSH-style fall; gloves with a leather palm (cycling or motorcycle) preserve the skin. Thin cycling gloves with a gel pad on the palm are the minimum. For off-road — full gloves with knuckle armor.

Footwear. Closed-toe shoes with a firm sole. Sandals and flip-flops are an anti-pattern: the foot slides off the deck during braking, and the toes are unprotected against a chance hit against a curb or another scooter. Sole rubber suited to dry and wet asphalt (not mountain trail), thickness ≥ 10 mm, with ankle support for long rides.

10. 8-point gear checklist

Helmet — non-negotiable. EN 1078 / CPSC for 25 km/h; NTA 8776 for 25–45 km/h; ECE 22.06 for 45+. Open the Virginia Tech STAR rating and pick a 4- or 5-star model.
MIPS / WaveCel / equivalent. Rotational mitigation reduces concussion risk by up to 50 % — costs $20–50 more, the best payback per dollar in all of safety gear.
Two-finger above brow + Y-junction + chin strap. A helmet that sits wrong works worse; check fit every time.
Wrist guards. The second-most critical unit after the helmet. 50–60 % reduction in FOOSH fracture risk. Splint minimum 5–6 cm above the wrist.
EN 1621-1 Level 1 knee + elbow. Cobbles, tram tracks, curbs — city plus light off-road. Level 2 — for off-road or 45+ km/h.
EN 1621-2 back protector. Optional for city; mandatory for off-road and hyperscooter. Replaces backpack-as-armor.
Eyewear ANSI Z87.1 / EN 166, gloves with leather palm, closed-toe shoes. The third tier — cheap and often ignored.
Replace after crash. Any impact = helmet replaced. Otherwise 3–5 years (CPSC) up to 5–10 years (Snell), depending on UV exposure and storage. Wear indicators: brittleness of the outer shell, cracking straps, foam that feels hard to the touch.