Design of a Nature-Inspired Honeycomb Bicycle Helmet for

Prevention of Traumatic Brain Injury

A. R. A. King1, A. Tyedmers1, S. Gonder1, B. Yu1, and C. E. Quenneville1
1
McMaster University

ABSTRACT

Following a head impact, EPS foam bicycle helmets effectively mitigate translational kinematics
(linked to skull fracture) but not rotational kinematics (associated with traumatic brain injury
(TBI)). Honeycomb structures can potentially be an advantageous alternative to foam due to their
superior energy absorption potential and anisotropic properties. Specifically, honeycomb allows
deformation in the shear loading direction to absorb impact energy that causes rotational head
kinematics. In beeswax, pentagons and heptagons (called ‘5-7 defects’) are observed within the
array of hexagons in areas of curvature. In this study, a new helmet design was proposed that uses
hexagonal honeycomb with 5-7 defects to accommodate the curvature of the human head. Out-of-
plane compressive properties of the design (relevant to skull fracture protection) were explored.
The optimal design would maximize energy absorption per unit volume (𝑈! ) while maintaining
peak stress (𝜎" ) just below the injury threshold for skull fracture. Various honeycomb designs,
including regular honeycomb (with six variations in relative density (𝜌̅ )), and honeycomb with 5-
7 defects were 3D-printed with thermoplastic polyurethane. Quasi-static (strain rate of 10-3 /s)
out-of-plane compression tests were performed. There were no notable differences in mechanical
properties between samples with and without 5-7 defects, indicating that 5-7 defects do not reduce
the energy absorbing capability. Increasing 𝜌̅ from 14 to 28% resulted in a linear increase in both
𝜎" and 𝑈! . Using an injury threshold for skull fracture of 2.25 MPa, the maximum allowable
honeycomb 𝜌̅ was determined to be 36%. Compression tests were also performed at strain rates
of 0.03 /s and 0.06 /s and the results indicated there was no strain rate dependence over this range.
The present work demonstrates the design is capable of protecting against skull fracture. This is
the first stage towards determining the optimal honeycomb design for head protection, where
future work will focus on reducing the risk of TBI.

INTRODUCTION

Traumatic brain injury (TBI) is a prevalent cause of disability and even death, contributing
to 30% of all injury deaths (Faul et al., 2010). Bicycle accidents are the leading cause of head
injuries from sports (AANS, 2014), with 81,000 individuals being treated in U.S. emergency
rooms in 2015 (CPSC, 2015). Of these head injuries, 13,000 involved diagnoses of TBI, such as
concussion (CPSC, 2015), and the incidence is likely even higher since less-severe head injuries
are often not treated in emergency rooms.

1
2022 The Ohio State University Injury Biomechanics Symposium
*This paper has not been peer-reviewed
 Traditional bicycle helmets are made of an expanded polystyrene (EPS) foam liner, with a
hard outer shell. They have been specifically designed according to the mandatory impact test
standard (CPSC, 1998) to mitigate linear kinematics of the head following simulated radial
impacts. Consequently, EPS foam helmets effectively reduce the risk of skull fracture, penetrating
injury, and severe brain injury (Cripton et al., 2014; Fahlstedt et al., 2014; McIntosh et al., 2013).
In contrast to standard impact tests, head impacts in real-world bicycle accidents typically occur
at an angle of 30 to 60 degrees with respect to the horizontal impact surface (Bourdet et al., 2012;
Bourdet et al., 2014). These ‘oblique’ impacts transmit an additional tangential force to the head
which causes rotational kinematics (McIntosh et al., 2013). Research has shown that the primary
mechanism of diffuse brain injuries, such as mild TBI or concussion, is rotational head kinematics,
which is measured by rotational velocity or acceleration (Holbourn et al., 1943; Gennarelli et al.,
1993; Gutierrez et al., 2001; Ommaya et al., 2002; Post et al., 2015). Studies have shown that
wearing an EPS foam helmet does not reduce the risk of sustaining mild TBI (Thompson et al.,
1989; Bambach et al., 2013). Therefore, advanced helmet designs must specifically target
mitigation of rotational kinematics by dampening the tangential force component of an oblique
impact.

Recently, bicycle helmets with dedicated rotation-damping systems have been introduced.
The most common rotation-damping system is the Multidirectional Impact Protection System
(MIPS) slip liner (MIPS AB, Täby, Sweden). MIPS employs a thin polycarbonate plastic liner
recessed within a traditional EPS foam helmet. Its purpose is to reduce the rotational acceleration
of the head upon impact by permitting 10-15 mm of movement between the helmet and the head.
A study that performed impact testing on helmets with MIPS slip liners showed a 22% reduction
in rotational acceleration of the head compared to traditional EPS foam helmets (Bliven et al.,
2019). Helmets that incorporate alternate cellular solids have also been introduced, such as
WaveCel (WaveCel, Wilsonville, USA). This helmet contains EPS foam with an additional
collapsible cellular solid layer. Rotational acceleration is reduced by flexing/bending of the cell
walls and gliding of the material relative to the EPS foam layer. A study that performed impact
testing on helmets with WaveCel technology showed a 73% reduction in rotational acceleration of
the head compared to traditional EPS foam helmets (Bliven et al., 2019).

The limitations of traditional helmets to protect the head during impact are attributed to the
mechanical response EPS foam. EPS foam does not sufficiently deform under shear loading to
address rotational accelerations (Di Landro et al., 2002). As well, EPS foam absorbs impact energy
through crack propagation, making it ‘single-use’ but despite this, people rarely discard their
equipment after an impact. A helmet design capable of providing protection for multiple impact
events would therefore reduce the frequency of head injuries. Due to the limitations of EPS foam
it is hypothesized that a design with an alternate material altogether, unlike MIPS and WaveCel,
would be beneficial.

Honeycomb structures can potentially be an advantageous alternative to foam due to their

superior energy absorption potential and anisotropic properties. Specifically, honeycomb allows
deformation in the shear loading direction to absorb impact energy that causes rotational head
kinematics. However, hexagonal honeycomb cannot cover a curved surface. In solving this
problem, inspiration was taken from beeswax honeycomb that often has complex external
2
2022 The Ohio State University Injury Biomechanics Symposium
*This paper has not been peer-reviewed
geometry involving curvature. These structures frequently contain pentagons and heptagons within
the array of hexagons, and these are known as ‘5-7 defects’. Several studies have demonstrated
that 5-7 defects can be intentionally included in a variety of arrangements for applications
involving curvature. For example, one study explored different arrangements including ‘stacked’
5-7 defects and ‘staggered’ 5-7 defects, where these two methods promoted different collapse
mechanisms in the neighboring hexagonal cells (Yu et al., 2022).

In this research, a new helmet design was proposed that uses hexagonal honeycomb with
5-7 defects to accommodate the curvature of the human head (Figure 1).

Figure 1: Full-scale helmet model (designed in nTopology (nTopology, New York, USA)) made of
hexagonal honeycomb with 5-7 defects (pentagons are outlined in red, and heptagons in blue).

To study the effectiveness of this proposed design for head protection in bicycle accidents,
two loading directions must be considered: out-of-plane compression (applicable to skull fracture
protection); and out-of-plane shear (applicable to diffuse TBI protection) (Figure 2).

Figure 2: a) Out-of-plane compressive loading direction on honeycomb (relevant to skull fracture

protection), b) Out-of-plane shear loading direction on honeycomb (relevant to TBI protection).

In the current study, out-of-plane compressive properties of the proposed design were
explored to find the optimal honeycomb design for skull fracture protection. This was the first
stage towards determining an optimal helmet design, where the second phase will focus on out-of-
3
2022 The Ohio State University Injury Biomechanics Symposium
*This paper has not been peer-reviewed
plane shear loading (relevant to TBI protection). The first design requirement was that the helmet
must absorb the kinetic energy of the head at impact through the potential energy of deformation
(strain energy). The second requirement was that the stress experienced by the head (𝜎" ) should
not exceed a specified limit corresponding to the injury threshold for skull fracture (𝜎#$%&%#'( ),
reported in the literature to be 2.25 MPa (Weisenbach et al., 2018). Based on the above
considerations, the optimal helmet design maximizes energy absorption per unit volume (𝑈! ) while
keeping peak stress (𝜎" ) just below the safety threshold. These parameters are shown on a typical
stress-strain curve for honeycomb under out-of-plane compression in Figure 3.

Figure 3: Typical stress-strain curve for honeycomb under out-of-plane compression. The stress-
strain graph has three primary regimes: linear elastic, plateau, and densification. The parameters
of interest include densification strain (𝜀) ), peak stress (𝜎" ), and energy absorption per unit
volume (𝑈! ). 𝑈! is given by the area under the curve before densification.

METHODS

Various honeycomb designs were generated in nTopology (nTopology, New York, USA),
including regular hexagonal honeycomb with six variations in relative density, honeycomb with
‘stacked’ 5-7 defects, and honeycomb with ‘staggered’ 5-7 defects (Figure 4). The size of the
samples (9.4 x 9.4 cm) was representative of the average contact area in a helmeted bicycle
accident (Bland et al., 2020), and the thickness of the samples (2.5 cm) was equal to that of a
typical EPS foam bicycle helmet. 3D-printing was employed for rapid prototyping of all samples
(one to three per design variation) with thermoplastic polyurethane filament, called Ninjatek
Cheetah (Ninjatek, Manheim, USA), an elastomeric material.

4
2022 The Ohio State University Injury Biomechanics Symposium
*This paper has not been peer-reviewed
 Figure 4: Top-view images of 3D-printed honeycomb models: regular honeycomb, honeycomb
with stacked 5-7 defects, and honeycomb with staggered 5-7 defects (pentagons are outlined in
red, and heptagons in blue).

Out-of-plane compression tests (Figure 2a) were performed using an Instron materials
testing machine to obtain stress-strain data. The test setup is shown in Figure 5. The strain rate (𝜀̇)
was quasi-static (10-3 /s), corresponding to a deformation rate of 1.5 mm/min and the samples were
compressed until after the onset of densification (Figure 3).

Figure 5: Out-of-plane compression test setup. Tests were performed using an Intron materials
testing machine with a 30 kN load cell.

The effects of 5-7 defect arrangement and relative density (𝜌̅ ) on 𝑈! and 𝜎" were
determined. 𝑈! was determined by computing the approximate integral of stress with respect to
strain, prior to densification (𝜀) ). 𝜀) was defined as the strain at which the slope of the tangent is
5
2022 The Ohio State University Injury Biomechanics Symposium
*This paper has not been peer-reviewed
equal to the slope of the elastic regime (Chan et al., 2003) (Figure 3). The reusability of the design
was evaluated by performing multiple compression cycles on each sample. The strain rate
dependence of honeycomb was then determined by comparing 𝑈! and 𝜎" between out-of-plane
compression tests performed at strain rates of 10-3 /s, 3x10-2 /s, 6x10-2 /s (two samples each).

RESULTS

Honeycomb specimens with stacked and staggered 5-7 defects were successfully 3D-
printed with a 𝜌̅ of 17 ±1% (two samples per design). There were no observable differences in
mechanical properties between samples with and without 5-7 defects (Figure 6 and Table 1).

Figure 6: Stress-strain responses for regular honeycomb, honeycomb with stacked 5-7 defects,
and honeycomb with staggered 5-7 defects (shaded region shows confidence interval).

Table 1: Results of compression tests for regular honeycomb and honeycomb with 5-7 defects
with percent differences from regular honeycomb.
Peak Stress Energy Absorption per
Design % difference % difference
(𝝈𝒑 ) Unit Volume (𝑼𝒗 )
Regular
0.697 MPa - 0.353 MPa -
Honeycomb
Honeycomb
with Stacked 0.718 MPa 2.9 % 0.377 MPa 6.6%
5-7 Defects
Honeycomb
with Staggered 0.603 MPa 13.5 % 0.319 MPa 9.7%
5-7 Defects

6
2022 The Ohio State University Injury Biomechanics Symposium
*This paper has not been peer-reviewed
 Regular honeycomb specimens were successfully 3D-printed with six variations in 𝜌̅ ,
ranging from 14% to 28% (one to three samples per design). Average out-of-plane compressive
stress-strain responses of designs with variation in 𝜌̅ are plotted in Figure 7.

Figure 7: Stress-strain responses for regular honeycomb with variation in 𝜌̅ (shaded region shows
confidence interval).

The effects of 𝜌̅ on 𝜎" and 𝑈! are illustrated in Figure 8 and Figure 9, respectively.
Increasing 𝜌̅ from 14 to 28% resulted in a 6-fold linear increase in 𝜎" (from 0.28 to 1.85 MPa) and
a 4-fold linear increase in 𝑈! (from 0.13 to 0.65 kJ/mm3). Using an injury threshold for skull
fracture (𝜎#$%&%#'( ) of 2.25 MPa (Weisenbach et al., 2018), the maximum allowable honeycomb 𝜌̅
was 36%.

7
2022 The Ohio State University Injury Biomechanics Symposium
*This paper has not been peer-reviewed
 Figure 8: Effect of 𝜌̅ on 𝜎" . The red dots are the result for each design variation, the red dashed
line is from (Weisenbach et al., 2018), and the black dotted line is a linear regression used to
extrapolate to the maximum allowable 𝜌̅ .

Figure 9: Effect of 𝜌̅ on 𝑈! . The blue dots give the result for each design variation, and the black
dotted line is a linear regression.

The stress-strain results for regular honeycomb with 𝜌̅ of 16 ± 0.3% under out-of-plane
compressive loading with three different strain rates are shown in Figure 10. There were no notable
differences in mechanical properties of the honeycomb under the various loading conditions.

8
2022 The Ohio State University Injury Biomechanics Symposium
*This paper has not been peer-reviewed
 Figure 10: Stress-strain responses for regular honeycomb with variation in compression test
strain rate (shaded region shows confidence interval).

Finally, upon testing the reusability of the design, the honeycomb structures experienced a
50% decrease in strength in the second cycle, due to prior cell wall bending.

DISCUSSION

A novel bicycle helmet design was proposed using hexagonal honeycomb with pentagon-
heptagon (5-7) defects. Out-of-plane compression tests were performed, with the primary
outcomes being energy absorption per unit volume (𝑈! ) and peak stress (𝜎" ). 𝑈! is an important
parameter because it is a measure of how much kinetic energy of the impact can be absorbed by
the material. Impact energy that is not absorbed by the protective device is transferred to the head,
causing high levels of stress and subsequent translational and rotational head kinematics. 𝜎" is
important in helmet design because it corresponds to the stress transmitted to the head upon impact
and if this value is too high, the helmet itself can directly cause injury to the head.

The effects of stacked and staggered 5-7 defects on the honeycomb mechanical properties
were explored. These two designs were chosen because of previous work that showed they
promoted different collapse mechanisms and therefore altered the strength of the material (Yu et
al., 2022). However, results of out-of-plane compression testing in the current study, related to
skull fracture protection, suggested that the inclusion of 5-7 defects had no effect on the mechanical
properties. Given that 5-7 defects must be included to accommodate the curvature of the head, this
result is beneficial in that the optimized design can be based on that for regular honeycomb.
However, a larger sample size is required to be confident in this null result.

𝜎" increased linearly in relation to 𝜌̅ , indicating the honeycomb cell walls inhibited plastic
collapse (Gibson et al., 2014). Consistent with this finding, the honeycombs did not have consistent
9
2022 The Ohio State University Injury Biomechanics Symposium
*This paper has not been peer-reviewed
mechanical properties over repeated tests. In high severity impacts (that can cause skull fracture),
energy absorption due to plastic deformation is beneficial; however, because helmets are rarely
discarded after moderate impacts (that may cause TBI), it is desirable to achieve elastic
deformation under shear loading. The recoverability of honeycomb under shear loading will be
explored in future work.

There was a linear increase in 𝑈! in relation to 𝜌̅ in the tested range. It is expected that with
a wider range of 𝜌̅ , 𝑈! would reach a maximum and then decrease to approach the 𝑈! of the solid
material. Future work will test this hypothesis. The relationship between 𝜎" and 𝜌̅ was extrapolated
to determine a maximum allowable 𝜌̅ of 36%, according to the threshold for skull fracture. In
extrapolating the observed 𝑈! relationship, honeycomb with 36% 𝜌̅ would be optimal. However,
the value of 𝜌̅ at which 𝑈! begins to decrease must be determined to be confident in this result.
Also, it is important to note that the skull fracture threshold of 2.25 MPa used in this study does
not represent the entire population, since injury thresholds vary with factors such as age and sex.

A limitation of the current study is that mechanical testing was primarily performed at a
quasi-static strain rate, which is not representative of a realistic head impact in a bicycle accident.
Therefore, the strain rate dependence of honeycomb was determined using strain rates up to 6x10-
2
/s, which was limited by the capabilities of the materials testing machine used. The strain rate
representative of a real-world head impact is approximately 250 /s. Although the results thus far
show there is no strain rate dependence, a test setup that allows a more realistic impact velocity
must be used in future work. Previous studies show that other cellular solids (foam) had no strain
rate dependence when testing at strain rates up to 200 /s (Liu et al., 2020; Rahimidehgolan et al.,
2022).

The initial findings suggest that honeycomb geometry can be modified to develop a design
optimized for head protection. In the next phase of this work, the effect of honeycomb geometry
on the out-of-plane shear properties, associated with TBI protection, will be determined.
Honeycomb is known to be anisotropic, therefore it is expected that sufficient deformation is
achievable in the shear loading direction to offer mitigation of rotational kinematics. The collective
results will be used to achieve the optimal honeycomb design, and this will be applied to a full-
scale helmet (Figure 1).

CONCLUSIONS

This study proposed a new helmet design to address the limitations of current EPS foam
helmets. The honeycomb helmet design demonstrates potential to effectively protect against both
skull fracture and TBI, by mitigating the translational and rotational kinematics of the head
following impact.

10
2022 The Ohio State University Injury Biomechanics Symposium
*This paper has not been peer-reviewed
 REFERENCES

AANS. (2014). Sports-Related Head Injury. American Association of Neurological Surgeons,

Rolling Meadows, IL.

BAMBACH, M.R., MITCHELL, R., GRZEBIETA, R.H., OLIVIER, J. (2013). The effectiveness
of helmets in bicycle collisions with motor vehicles: A case–control study. Accident
Analysis and Prevention, 53, 78–88.

BLAND, M.L., MCNALLY, C., CICCHINO, J.B., ZUBY, D.S., MUELLER, B.C.,
MCCARTHY, M.L., NEWGARD, C.D., KULIE, P.E., ARNOLD B.N., ROWSON S.
(2020). Laboratory Reconstructions of Bicycle Helmet Damage: Investigation of Cyclist
Head Impacts Using Oblique Impacts and Computed Tomography. Annals of Biomedical
Engineering, 48(12), 2783-2795.

BLIVEN, E., ROUHIER, A., TSAI, S., WILLINGER, R., BOURDET, N., DECK, C.,
BOTTLANG, M. (2019). Evaluation of a novel bicycle helmet concept in oblique impact
testing. Accident Analysis and Prevention, 124, 58-65.

BOURDET, N., DECK, C., CARREIRA, R.P., WILLINGER, R. (2012). Head impact conditions
in the case of cyclist falls. Journal of Sports Engineering Technology, 226(3/4), 282–289.

BOURDET, N., DECK C., SERRE, T., PERRIN, C., LLARI, M., WILLINGER, R. (2014). In-
depth real-world bicycle accident reconstruction. International Journal of Crashworthiness,
19(3), 222–232.

CHAN, K.C., XIE, L.S. (2003). Dependency of Densification Properties on Cell Topology of
Metal Foams, Scripta Materialia, 48(8), 1147–1152.

CPSC. (2015). National Electronic Injury Surveillance System Database. United States Consumer
Product Safety Commission, Rockville, MD.

CPSC. (1998). Safety standard for bicycle helmets final rule (16 CFR Part 1203). Rockville, MD:
United States Consumer Product Safety Comission, 11711–11747.

CRIPTON, P.A., DRESSLER, D.M., STUART, C.A., DENNISON, C.R., RICHARDS, D..
(2014). Bicycle helmets are highly effective at preventing head injury during head impact:
head-form accelerations and injury criteria for helmeted and unhelmeted impacts. Accident
Analysis and Prevention, 70, 1–7.

DI LANDRO, L., SALA, G., OLIVIERI, D. (2002). Deformation mechanisms and energy
absorption of polystyrene foams for protective helmets. Polymer Testing, 21(2), 217-228.

FAUL, M., XU, L., WALD, M.M., CORONADO, V. (2010). Traumatic brain injury in the United
States. Atlanta, GA: Centers for Disease Control and Prevention, National Center for Injury
Prevention and Control.
11
2022 The Ohio State University Injury Biomechanics Symposium
*This paper has not been peer-reviewed
FAHLSTEDT, M., HALLDIN, P., KLEIVEN, S. (2014). Importance of the bicycle helmet design
and material for the outcome in bicycle accidents. In: Paper presented at: International
Cycling Cycling Safety Conference.

GENNARELLI, T.A. (1993). Mechanisms of brain injury. Journal of Emergency Medicine,

11(Suppl 1), 5–11.

GIBSON, L.J., ASHBY, M.F. (2014). Cellular solids: Structure and properties, second edition.
Cambridge University Press.

GUTIERREZ, E., HUANG, Y., HAGLID, K. (2001). A new model for diffuse brain injury by
rotational acceleration: I model, gross appearance, and astrocytosis. Journal of
Neurotrauma, 18(3), 247–257.

HOLBOURN, A.H. (1943). Mechanics of head injuries. Lancet 2, 438– 441.

LIU, Y., RAHIMIDEHGOLAN, F., ALTENHOF, W. (2020). Anisotropic compressive behavior

of rigid PVC foam at strain rates up to 200 s− 1. Polymer Testing, 91, 106836.

MCINTOSH, A.S., LAI, A., SCHILTER, E. (2013). Bicycle helmets: head impact dynamics in
helmeted and unhelmeted oblique impact tests. Traffic Injury Prevention, 14(5), 501–508.

OMMAYA, A.K., GOLDSMITH, W., THIBAULT, L.. (2002). Biomechanics and

neuropathology of adult and paediatric head injury. British Journal of Neurosurgery, 16(3),
220–242.

POST, A., THOSHIZAKI, T.B. (2015). Rotational acceleration, brain tissue strain, and the
relationship to concussion. Journal of Biomechanical Engineering, 137(3),030801.

RAHIMIDEHGOLAN, F., LIU, Y., ALTENHOF, W. (2022). Experimental, numerical and

analytical investigations on the elevated strain rate compressive behavior of high-
performance PES foam up to 200 s− 1. International Journal of Impact Engineering, 161,
104088.

THOMPSON, R.S., RIVARA, F.P., THOMPSON, D.C. (1989). A case-control study of the
effectiveness of bicycle safety helmets. The New England Journal of Medicine, 320, 1361–
1367.

WEISENBACH, C.A., LOGSDON, K., SALZAR, R.S., CHANCEY, V.C., BROZOSKI, F.

(2018). Preliminary investigation of skull fracture patterns using an impactor
representative of helmet back-face deformation. Military medicine, 183(Suppl 1), 287-93.

YU, B., VAN EGMOND, D.A., ABU SAMK, K., ERBB, U., WILKINSON, D., BRÉCHET, Y.,
EMBURY, D., ZUROB, H. (2022). The Design of “Grain Boundary Engineered” Cellular
Materials: The Role of 5-7 Defects in Hexagonal Honeycomb. [Unpublished manuscript].
12
2022 The Ohio State University Injury Biomechanics Symposium
*This paper has not been peer-reviewed