Content uploaded by David J Pearsall
Author content
All content in this area was uploaded by David J Pearsall
Content may be subject to copyright.
© 2007
isea
Sports Engineering (2007) 10, 65–74 65
Introduction
In addition to skating and stick skills, ice hockey
involves numerous collisions between players and the
environment; hence, the game involves an inherent
risk of injury. In leagues permitting body checking,
previous epidemiological studies have shown that the
head and neck have consistently ranked as two of the
most commonly injured body sites, followed by the
knee, shoulder, hip/thigh/groin, and hands/fingers
(Azuelos et al., 2004). More specific studies on head
injuries in ice hockey reported that concussions (or
mild traumatic brain injuries, mTBI) account for
between 2% and 14% of all hockey-related injuries,
and between 15% and 30% of all hockey-related head
injuries (Goodman et al., 2001). More recently, Flik
and colleagues (2005) reported that head injuries in ice
hockey represent up to 20% of all trauma. In short,
concussions form a substantial and persistent problem
in ice hockey that has acute and long-term conse-
quences for the health of the players (Biasca et al.,
2002; Flik et al., 2005; Goodman et al., 2001).
Correspondence address:
David Pearsall
Department of Kinesiology & Physical Education
McGill University, 475 Pine Avenue West
Montréal, Québec, Canada H2W 1S4
Tel: 001 514 398 4184, extension 09976
Fax: 001 514 398 4186
E-mail: david.pearsall@mcgill.ca
Evaluation of impact attenuation of facial protectors
in ice hockey helmets
M. Lemair and D.J. Pearsall
Department of Kinesiology & Physical Education, McGill University, Montréal, Québec, Canada
Abstract
The purpose of this study was to investigate the extent to which ice hockey facial protectors can
decrease overall head acceleration during blunt impacts, as well as to identify whether attenua-
tion differences exist between visors and cages. Commercial models of three cages and three
visors were assessed. Blunt impacts were simulated, permitting the measurement of peak acceler-
ations (PA) within the surrogate headform. Results indicated that face protectors, in combination
with helmets, substantially reduced PA during blunt impacts within threshold safety limits
(below 300g). In general, cages showed lower PA than visors. Differences between models were
also observed during repeated impacts and impact site. In conclusion, this study demonstrates
that facial protectors function beyond their role in solely preventing facial injuries, complement-
ing the role of the helmet in attenuating head deceleration during impact. Consequently, the
utilisation of facial protectors may reduce the severity and incidence of head injuries.
Keywords: helmets, impact, facial protection, ice hockey
10.2.1 Sports F73 30/8/07 16:48 Page 65
To address this specific concern, the wearing of
helmets (covering the cranial skull of a player’s head) is
obligatory in competitive contact ice hockey leagues.
Helmets are designed to diminish the magnitude of
impact forces during collisions by distributing the
contact load over a wider area of the cranium and by
means of an energy-absorbing liner system.
Manufactured helmets must pass safety standard
guidelines such as ASTM International F1045-04 or
NOCSAE DOC 001-04m05 standards for certifica-
tion. Typically, these standardised tests involve
controlled-impact simulations of surrogate headforms
with helmets to assess impact attenuation gains (as
quantified by reduced peak headform acceleration). In
most contact sports, the intervention of helmets has
reduced the severity of head injuries.
The potential to reduce concussion incidence by
wearing facial protectors in conjunction with helmets
has been proposed (Biasca et al., 2002). Some prospec-
tive cohort studies have been conducted to investigate
this possibility. For instance, Benson et al. (1999)
monitored the number of head or neck injuries
sustained among intercollegiate players wearing full-
face shields compared to those wearing half-face
shields. They found no evidence to support specula-
tion that the full-face shield increases a player’s risk of
sustaining a neck injury or concussion. In a later study,
Benson et al. (2002) reported on concussion incidence
rates in ice hockey. Independent of prior injury
history, position played and experience, the half-face
shield players lost a greater amount of times than full-
face shield players (average of 3.29 sessions per
concussion versus 1.70, respectively). In a similar
prospective study by Stuart and colleagues (2002),
athletes were monitored within competitive junior A
league where none, partial and full facial protection
was allowed. Stuart et al. observed that players wearing
no facial protection were injured at a rate more than
twice that of players wearing partial protection and
almost seven times more than players wearing full
facial protection. Therefore, full-face guards were
found to provide almost a 5-fold reduction of eye
injury incidences and a rate of concussion reduced
from 12.2 to 2.9 concussions per 1000 players’ hours,
compared to no facial protection. Similarly, in a review
of concussion frequency in the NHL during the
2001–2002 season (Stevens et al., 2006), the use of
visors was not shown to significantly affect the preva-
lence of concussion, though other head and face
injuries were minimised. Hence, these studies suggest
that the use of facial protectors decreases, at least, the
severity and, at most, the rate of head injuries.
The dynamics of impact at an incident vector
towards the face in terms of net head acceleration is
not well understood with respect to concussion events,
let alone the intervening effects of facial protectors in
combination with helmets. Thus, the purpose of this
project was to determine:
1 if facial protectors can attenuate head acceleration
within acceptable limits (i.e. below 300g) during
blunt facial collisions
2 if acceleration attenuation differs between cages
and visor facial protectors conjoined with ice
hockey helmets
3 if helmet liner materials (vinyl nitryl, VN, or
expanded polypropylene, EPP) significantly modify
the above responses.
Materials and methods
Materials
Six models of commercial ice hockey facial protectors
were evaluated: three full-face shields (cages) and
three half-face shields (visors) (Fig. 1). The cage
models included the Bauer Nike FM8500, Itech FM
480 and the CCM RBE VIII. The visor models
included the Bauer Nike FM1000, Itech HLC and the
Oakley Aviator. Four samples of each model were
tested for each condition as described below (see
‘Procedure’). During testing, all facial protectors were
mounted on a common reference ice hockey helmet
model (medium NBH8500). In addition to the above,
two separate liners covering the inside of the ice
hockey helmet were evaluated: VN and EPP foams.
A drop rig monorail guide assembly (specifications
according ASTM F1045-04 standards; Fig. 2) was
used to control the free drop height and direction of
the headform mounted with the helmet and facial
protector. A full-facial NOCSAE headform (DOC
001-04m05 standard) was attached to the armature of
the drop rig monorail guide. The headform’s orienta-
tion was adjustable, allowing impacts to be delivered
to any point on the helmet or the facial protector.
66 Sports Engineering (2007) 10, 65–74 © 2007
isea
Impact attenuation of facial protectors in ice hockey helmets M. Lemair and D.J. Pearsall
10.2.1 Sports F73 30/8/07 16:48 Page 66
© 2007
isea
Sports Engineering (2007) 10, 65–74 67
M. Lemair and D.J. Pearsall Impact attenuation of facial protectors in ice hockey helmets
Figure 1 Models of facial protectors evaluated
(a) NBH FM8055 (d) Oakley Aviator
(b) CCM RBE VIII (e) NBH FM1000
(c) Itech FM480 (f) Itech HCL
10.2.1 Sports F73 30/8/07 16:48 Page 67
68 Sports Engineering (2007) 10, 65–74 © 2007
isea
Impact attenuation of facial protectors in ice hockey helmets M. Lemair and D.J. Pearsall
Headforms with approximate facial geometry were
essential to assess the cages and visors, since the latter
are designed to function around the former. The
medium-size headform (head circumference of
578 mm) was used during impact testing. The
combined mass of the assembly (headform and
armature) was 6.98 kg. Within the headform, a tri-
axial accelerometer (±500g) was mounted at its centre
of mass. An analogue/digital data acquisition card
(Cadex Inc. version 4.6) was used to record and
condition the output acceleration signals (frequency
range 0–1000 Hz with a ±1.5% variance; sampling
rate 10 kHz; filtering acquisition use 1000 Hz). From
this data, peak acceleration (PA) and Gadd Severity
Index (GSI) scores could be obtained from each
impact. All tests were conducted at ambient room
temperature (20 ± 2°C) and relative humidity of 55%.
The impact surface was a flat steel anvil that was
supported on a rigid, concrete foundation.
Figure 2 Monorial drop rig tester; headform and steel anvil.
Procedure
The testing protocol followed the ASTM F1045-04
standard using the NOCSAE headform. Facial pro-
tectors and helmets were positioned on the headform
(attached to the armature and guide rail) then dropped
77 ± 1 cm in order to achieve a pre-impact kinetic
energy of 45 ± 3 J. During impact, acceleration
measures were recorded over a 20 ms window.
Helmet positioning and tightness were checked
before each impact test. The helmet’s ear aperture was
aligned concentrically with the headform’s index ear
holes and the helmet’s front rim was positioned along
a pre-measured line on the headform forehead (5 cm
above the basic plane). By re-orienting the headform
on the drop rig’s armature, each helmet and facial
protector were impacted at four different sites (Fig. 3)
including:
1 the crown (C) of helmet (the intersection of the
mid-sagittal and the coronal planes); the crown was
used as a reference impact site to compare direct
helmet to direct facial protector impacts;
2 the front (F) of the facial protector (located at the
intersection of the basic plane and the coronal plane);
3 at 45º to the front boss (FB45) of the facial protector
(located on the basic plane at a 45° angle from the
coronal plane); impact was on the left cheek; and,
4 the J clip (JCL) of the helmet (located on the basic
plane at the approximate intersection of the basic
plane and the mid-sagittal plane). The J clip
(shaped in the form of a ‘J’) secures the facial
protector to the side of the helmet. Impacts were
done on the left ear side of the headform.
Before impact, the headform was oriented to impact
the designated site. Then, the helmet and the facial
protector were mounted on the headform. Therefore,
impacts were conducted with respect to the reference
lines on the headform. Prior to each drop test, the
helmet was adjusted on the headform. Each site was
impacted three times with 1-minute intervals between
tests. Both new samples of helmet and facial protec-
tors were used for each site-specific test, so as to avoid
potential cumulative damage effect.
In addition to the above, prior to each test the face
of the headform was covered with a 1 mm-thick
coating of white contact paste. After impacts, if face
10.2.1 Sports F73 30/8/07 16:48 Page 68
© 2007
isea
Sports Engineering (2007) 10, 65–74 69
contact occurred, the extent and location of contact
could be identified by paste marks transferred either
to the facial protector and/or the steel anvil. Finally, as
a visual log of collisions, F impacts were filmed using a
high-speed video camera (TroubleShooter, Fastec
Imaging Inc.) at 1000 frames s–1. The camera was
positioned lateral and perpendicular to the monorail’s
drop plane. Subsequent analysis permitted the linear
kinematics of headform and facial protector to be cal-
culated. Specifically, the amount of deformation
(linear compression) of the headform’s facial structure
(nose) was measured. This was accomplished by
tracking the change in relative displacements between
the facial cheek and nose tip using motion measure-
ment software of the scaled digital images.
Data analysis
With a standard reference helmet (NBH8500) lined
with VN foam, peak acceleration (PA) was evaluated
with respect to the independent variables (model, site,
repeated impacts) by means of a 3-way ANOVA:
PA = model6×site3×repeat impact3
Three samples of models were tested. For two helmet
models, the effects of two different liner materials
(EPP, VN) were assessed using a 4-way ANOVA:
PA = model2×site3×repeat impact3×foam2
Five samples of models and liner foam were tested.
Statistical analyses were performed with Statistica
(6.0, GLM & Post Hoc tests). Significant differences
were evaluated at αlevel of 0.05. Given that the intent
is not to endorse nor denigrate particular products, in
the following results specific name brands will not be
cited. Instead, they will be referred to as cage A, B, C
and visor A, B, C (in no particular order).
Results
For direct helmet impacts (C site), PA was lower for
VN-lined ice hockey helmets than for EPP (Fig. 4,
p< 0.001). PA for VN liners ranged from 75gto 110g,
whereas PA for EPP liners ranged from 90gto 130g.
The wide range of PA was dependent on the repeated
impact sequence (p= 0.028). The general trend
observed was that the PA increased progressively from
the first to third impacts. Significant differences
between the first and the third impacts were calculated
(p= 0.036). However, for impacts direct at the facial
protectors (F, FB45, JCL), the type of helmet liner had
no significant effect (p= 0.39, Fig. 5).
Nonetheless, as will be shown in the following
results (irrespective of helmet liner), use of cages and
visors demonstrated substantial and significant reduc-
tions in PA. To put in perspective the magnitude of the
effects introduced, note that from several pre-testing F
impacts of the uncovered headform, PA values ranged
from 380gto 420g, whereas with the use of an ice
hockey helmet alone, PA were reduced to between
100gand 130g. Furthermore, with the attachment of a
face protector, PA were further reduced from 10gto
100g. The following results will provide specific details
on influence models, site and repeated impacts on PA.
M. Lemair and D.J. Pearsall Impact attenuation of facial protectors in ice hockey helmets
Figure 3 Impact location sites.
(a) Top view
(b) Side view
10.2.1 Sports F73 30/8/07 16:48 Page 69
70 Sports Engineering (2007) 10, 65–74 © 2007
isea
Impact attenuation of facial protectors in ice hockey helmets M. Lemair and D.J. Pearsall
Liner
VNEPP
*EPP, VN (p < 0.001)
*1st impact, 3rd impact (p = 0.036)
One
Two
Three
0
20
40
60
80
Peak acceleration (g)
100
120
140
*Cage A, Visor A (p = 0.009)
*Front, FB45 (p = 0.029)
*Front, JCL (p = 0.006)
Front
EPP VN
Visor A
Liner, facial protector
EPP VN
Cage A
FB45
JCL
0
20
40
60
80
Peak acceleration (g)
100
120
140
160
180
200
Cage A Cage B Cage C Visor A
Facial protector
Visor B Visor C
*Cage A, Cage C at 3rd impact (p<0.012)
One
Two
Three
0
20
–20
40
60
80
Peak acceleration (g)
100
120
140
Figure 4 Peak acceleration (g) for crown impacts at third impact.
Figure 5 Peak acceleration
(g) as a function of helmet
liner and impact site for
third impact.
Figure 6 Peak acceleration (g)
as a function of impact
number and front impact site.
10.2.1 Sports F73 30/8/07 16:48 Page 70
The type of facial protector model tested affected
PA (p= 0.004). In general, PA was lower for cages than
for visors (Figs. 5–6). There were no significant differ-
ences between the three cage models, but visor C
showed significantly lower PA than visors A and B
(p= 0.00072 and 0.00081, respectively).
By impact sites, PA was lowest at F and highest at
the JCL sites (Fig. 7). Significant differences were
noted between impact sites (F > FB45, p= 0.016;
F > JCL, p= 0.0001; JCL > FB45, p= 0.0001).
Repeated impacts significantly increased PA with each
consecutive test (i.e. 1st < 2nd, 1st < 3rd, p= 0.0001;
2nd < 3rd, p= 0.021).
Interactions were identified for specific models and
sites (Table 1); for instance, cages A and C were signif-
icantly different (p= 0.0128) at the third impact. In
addition, visor C was significantly different from
visors A and B (p= 0.0001 and p= 0.019) at FB45.
Cages A and C were significantly different (p= 0.013)
at F.
From the high-speed imaging, the nature and
extent of facial contact was considerably different
between the six facial protectors (Fig. 8). The amount
of resulting facial (nose) compression closely paral-
leled PA. In general, facial compression was greater
in visors than in cages (p< 0.001) and was dependent
© 2007
isea
Sports Engineering (2007) 10, 65–74 71
M. Lemair and D.J. Pearsall Impact attenuation of facial protectors in ice hockey helmets
Cage A Cage B Cage C Visor A
Facial protector
Visor B Visor C
*Cage A, Cage C at Front (p = 0.013)
Front
FB45
JCL
0
Peak acceleration (g)
300
250
200
150
100
50
*Visor C, Visor A at FB45 (p < 0.001)
*Visor C, Visor B at FB45 (p = 0.019)
Cage A Cage B Cage C Visor A
Facial protector
Visor B Visor C
*Cage B, Cage C (p < 0.001)
*Cage B, Cage A (p = 0.041)
0
Distance of nose compression (cm)
1.2
1
0.8
0.6
0.4
0.2
*Visor C, Visor A (p = 0.027)
**
Figure 8 Distance (cm) of nose
compression for facial protectors
for third impacts at front site.
Figure 7 Peak acceleration (g)
as a function of impact site for
third impact.
10.2.1 Sports F73 30/8/07 16:48 Page 71
on the model. The effect of repeated impacts was
more substantial for certain facial protector models.
For cages, the impact repetition significantly affected
the extent of nose compression (p< 0.001), whereas
for visors, there were no significant differences in nose
compression for each impact repetition (p= 0.307).
There were significant differences across each model.
Cage B was significantly less than cages A and C
(p= 0.041 and p< 0.001, respectively) and visor C was
significantly lower than visor A (p= 0.027).
During F impacts, the cages progressively deformed
with repeated impacts. Typically, the cages would not
return to their original spherical shape. Structural
deformation was restricted to the area of impact with
the wires bent inward, such that the impact point was
flattened. The extent and form of deformation varied
between cages but typically involved deformation over
2–3 wire rows (~5 cm). Conversely, with the visors, no
permanent deformation comparable to the flattening in
cages was observed. Instead, visors flexed with impact
and returned to their original form after impact (as
observed from visual digital recording at front impacts).
From post impact visual inspection, some fractures or
breakage were observed; for instance, the fracture of the
upper plastic support designed for aeration.
In addition to the above, the use of superficial paste
applied over the headform’s face indicated that during
F and FB45 impacts, contact occurred between the
face and the facial protector or anvil. For cages, paste
marks from the nose of the headform were left on the
steel anvil, increasing in size with repeated impact.
The chin support from the cage prevents any slippage
motion of the helmet and facial protector. Also, the
chin support prevents any contact of the mouth on the
steel anvil. For visors, paste marks were found both on
the inside of the visor, and from the mouth of the
headform onto the steel anvil. These also increased in
size with repeated impacts. Moreover, slippage of the
nose–visor contact point at impact was observed.
Consequently, the cage facial protector provides more
extensive contact shielding.
For FB45 impacts, cages show similar deformation
patterns as observed in front impacts, with perma-
nently bent wires and focal flattening of the cage. In
contrast to F impacts, the damage is not restricted to
the area of impact. The flattening on the right side
(impact side) of the cage causes a bulging or outward
buckling of the wires of the cage on the left side
(opposite impact site). Lack of paste marks on the
cheek indicated non-contact with the side of the cage
during impact. Presumably, the chin support prevents
the cheek from headform slippage.
Unlike the cages, the visors retake their original
shape but some damage remote to the impact point
was noted. Fissures within the plastic were observed
on the sides of the visor both at the lateral and medial
portions of the visor. Paste mark observations indicate
that the contact of the cheek with the visor progres-
sively increased with impact repetition. On the inside
of the visor, the surface area of the paste mark
gradually increased. Moreover, by third impact, there
was paste from the mandible side jaw on the steel
anvil. Thus, unlike the cages, the visor did not prevent
the chin from contacting the steel anvil.
For the most lateral site (JCL), minimal deforma-
tion to the cages and visors was shown probably to
72 Sports Engineering (2007) 10, 65–74 © 2007
isea
Impact attenuation of facial protectors in ice hockey helmets M. Lemair and D.J. Pearsall
Table 1 Average peak accelerations (a) (g± SD) and Gadd Severity Index (b) (g*s ± SD) measures for third impacts.
a)
g
(
g
±SD) Cage A Cage B Cage C Visor A Visor B Visor C
Site
FB45 53.6 ± 0.8 63.4 ± 18.2 69.5 ± 18.7 214.1 ± 115.2 165.0 ± 14.2 64.5 ± 9.2
Front 107.1 ± 5.4 87.1 ± 23.7 9.4 ± 2.6 107.8 ± 3.7 86.4 ± 3.1 88.9 ± 5.4
JCL 146.6 ± 22.3 86.4 ± 2.9 88.0 ± 4.3 157.6 ± 24.0 210.6 ± 54.9 212.9 ± 79.8
b) GSI (
g
*s ±SD) Cage A Cage B Cage C Visor A Visor B Visor C
Site
FB45 80.0 ± 35.4 135.0 ± 67.4 159.3 ± 72.7 558.5 ± 178.9 690.3 ± 153.7 180.0 ± 16.9
Front 357.5 ± 15.9 262.3 ± 105.0 0.5 ± 0.1 398.0 ± 7.3 242.0 ± 11.3 301.3 ± 40.1
JCL 508.0 ± 71.0 236.8 ± 13.1 264.8 ± 19.2 545.0 ± 97.6 784.3 ± 198.0 832.0 ± 364.0
10.2.1 Sports F73 30/8/07 16:48 Page 72
concurrent impact with the side of the helmet shell.
The JCL was not damaged, though scratching on the
interior and exterior portions of the facial protector was
seen. On the exterior, abrasion occurred due to slippage
on the steel anvil. On the interior, abrasion due to
contact with the J clip anchoring screws occurred.
Structural damage was only observed in cages where in
wires were deformed focally over 1 to 2 wire gaps.
Discussion
This study demonstrated that facial protectors can
serve to attenuate impact accelerations below accept-
able tolerance criteria (i.e. below 300g) as applied to
helmet shells and liners. Hence, it is appropriate to
conclude that the use of facial protectors in combina-
tion with helmets can further reduce the risk of head
injuries. This point is self-evident if one compares
peak acceleration measures with, and without, helmets
and protectors. From calibration testing, facial
impacts of the headform alone were as high as 400g–
well above the safety threshold tolerance. With the use
of an ice hockey helmet alone, PA decreased by up to
4-fold during front impacts. Though initially unex-
pected, the helmet shell covering the forehead was
observed to hit the outer portions of the steel anvil
concurrently with facial contact, thereby reducing the
PA for impacts directed along the basic plane. The use
of a facial protector combined with the ice hockey
helmet showed greater reductions in PA. Indeed, facial
protectors exhibited comparable PA attenuation to ice
hockey helmet shells and liners. Consequently, facial
protectors fulfill their primary function in providing
eye protection, while simultaneously reducing the PA
of direct impact of forces from facial vector incidence.
The types of helmet liner were shown to affect
impact response to the helmet (Fig. 4) but not to the
facial protector. It was speculated that the different
rigidity of the two foams tested may change the
stability of the facial protectors anchoring to the shell;
in turn altering the latter’s impact response. However,
for impacts on the facial protectors at all sites, no sig-
nificant differences were observed. Thus, facial
protectors may be considered to function independ-
ently of the liner.
Overall, cages perform better than visors in terms
of reducing PA (Figs. 4–6). There are several factors
that may be conjectured to explain such a result.
Primarily, the cages include a firm chin support,
whereas a visor’s lower margin is not supported and
easily collapses during impact by buckling inward and
downward. The chin support may have permitted
force distribution of the facial protector both above
and below the basic plane; whereas with the visor, the
forces can be distributed only above the basic plane.
Secondly, based on qualitative observations the cages
and visors exhibited two different behaviours upon
impact. Typically, cages were permanently deformed
focally at the area of impact including the bulging of
the wires and flattening of the cage, whereas visors
offered insufficient resistance to the forces of impact,
collapsing but then elastically rebounding to their
original geometric configuration. Lastly, the
geometric design and anchoring of the cage and the
visor differ substantially. The cage typically forms a
hemisphere covering the entire face, whereas the visor
forms a hemi-cylinder covering only the upper
portion of the face. Consequently, at impact, the
forces are distributed differently for the cage and the
visor. The cage shape may permit the radial distribu-
tion of the forces, whereas the visor shape may only
allow forces to be distributed medially-laterally
parallel to the basic plane.
A major factor in this research project was the effect
of repeated impact. A progressive increase in the PA
corresponding to a decrease in impact attenuation
capacity of the facial protector was observed. The
repeated impact effect was present for all facial
protector models and impact sites, indicating cumula-
tive structural damage and/or material degradation.
It was expected that impact sites would affect PA.
Given that the protector-to-face distance progres-
sively decreases laterally, a corresponding increase in
PA was expected. However, this general assumption
was not evident, such that each facial protector model
had different site-specific PA responses. Hence, other
design model specific parameters interact to alter PA
site behaviour.
The fact that facial contact occurred during
impact with the face protector in place should raise
some concern. Most safety standards require a non-
contact criterion for certification. For instance, to
obtain NOCSAE certification (i.e. NOCSAE DOC
(ND) 021-98m05a), facial protectors are submitted
© 2007
isea
Sports Engineering (2007) 10, 65–74 73
M. Lemair and D.J. Pearsall Impact attenuation of facial protectors in ice hockey helmets
10.2.1 Sports F73 30/8/07 16:48 Page 73
to standardised projectile tests wherein no impactor
contact to the face is accepted. In such tests, pucks are
propelled at the facial protector at 28 ms–1.
Considering the weight of the puck (120 g), this repre-
sents an impact of 65 J. From this study, high-speed
filming and paste marks clearly demonstrated that
facial contacts occurred from the drop test of 45 J.
However, given fundamental differences between the
projectile and drop tests, caution in exact comparison
is necessary. The projectile testing consists of a high
velocity, low mass impact to a small surface area,
whereas the drop test consists of a low velocity, high
mass impact on a large surface area. Furthermore,
facial contact injuries in the latter scenario are rare,
given epidemiological evidence (Stuart et al., 2002;
Benson et al., 2002).
During F impacts, the facial protectors as a whole
were both driven backwards and locally deformed,
permitting transient facial contact. Concurrent
forward slippage of the headform occurred. The
distance of facial (nose) compression varied with facial
protector models. As with PA, less compression
happened with cages than with visors. This is probably
due to the better support of the entire face in cages,
than in visors. For cages, compression progressively
increased with impact repetition up to an average of
8 mm maximum. For visors, compression was not
dependent on the impact repetition but was, in
general, larger (up to 1 cm on average). Further review
of the implications of contact permitted despite the
facial protector presented is recommended.
Conclusion
The prevalence of head injuries in ice hockey has been
reported in many epidemiological studies (Azuelos et
al., 2004; Goodman et al., 2001; Flik et al., 2005). This
study addressed the effect of facial protection on
impact attenuation. Facial protectors were found to
reduce PA transmitted to the headform, with cages
yielding significantly lower PA than visors. The
benefit of decreased PA from facially directed impacts
may in turn decrease the severity and/or incidence of
head injuries. This benefit may, in part, explain the
reduced head injuries reported in the respective
studies of Benson et al. (2002) and Stuart et al. (2002).
Acknowledgements
Assistance from Nike Bauer Hockey Inc. and financial
support from the Natural Science & Engineering
Research Council of Canada were greatly appreciated.
References
ASTM International (2004) Standard Performance
Specification for Ice Hockey Helmets, F1045-04.
Azuelos, Y.H., Pearsall, D.J., Turcotte, R. & Montgomery,
D.L. (2004) A review of ice hockey injuries: Location,
diagnosis, mechanism, in D.J. Pearsall & A.B. Ashare
(eds) Safety in Ice Hockey, American Society for Testing
and Materials, West Conshohocken.
Benson, B.W., Mohtadi, N.C.H., Rose, M.S. & Meeuwisse,
W.H. (1999) Head and neck injuries among ice hockey
players wearing full face shields vs half face shields.
Journal of American Medical Association, 282(24),
2328–2332.
Benson, B.W., Rose, M.S. & Meeuwisse, W.H. (2002) The
impact of face shield use on concussion in ice hockey: a
multivariate analysis. British Journal of Sports Medicine,
36, 27–32.
Biasca, N., Wirth, S. & Tegner, Y. (2002) The avoidability
of head and neck injuries in ice hockey: an historical
review. British Journal of Sports Medicine, 36, 410–427.
Flik, K., Lyman, S. & Marx, R.G. (2005) American
collegiate men’s ice hockey an analysis of injuries.
American Journal of Sports Medicine, 33(2), 183–187.
Goodman, D., Gaetz, M. & Meichenbaum, D. (2001)
Concussions in hockey: there is cause for concern.
Medicine & Science in Sports & Exercise, 33(12),
2004–2009.
NOCSAE (2005) Standard Projectile Impact Test Method
and Equipment Used in Evaluating the Performance
Characteristics Protective Headgear, Faceguard or
Projectiles. NOCSAE DOC (ND) 021-98m05a.
NOCSAE (2005) Standard Drop Test Method and
Equipment Used in Evaluating the Performance
Characteristics of Protective headgear. NOCSAE DOC
(ND) 001-04m05a.
Stevens, S.T., Lassonde, M., De Beaumont, L. & Keenan,
J.P. (2006) The effect of visors on head and facial injury
in national hockey league players. Journal of Science &
Medicine in Sport, 11(3), 238–242.
Stuart, M.J., Smith, A.M., Malo-Ortiguera, S.A., Fischer,
T.L. & Larson, D.R. (2002) A comparison of facial
protection and the incidence of head, neck, and facial
injuries in junior A hockey players. American Journal of
Sports Medicine, 30, 39–44.
74 Sports Engineering (2007) 10, 65–74 © 2007
isea
Impact attenuation of facial protectors in ice hockey helmets M. Lemair and D.J. Pearsall
10.2.1 Sports F73 30/8/07 16:48 Page 74
