ARTICLE IN PRESS

Journal of Biomechanics 41 (2008) 2972–2980

Contents lists available at ScienceDirect

journal homepage: www.elsevier.com/locate/jbiomech

Journal of Biomechanics

0021-92

doi:10.1

� Corr

E-m

www.JBiomech.com

Biomechanical response of the human mandible to impacts of the chin

Matthew Craig a, Cynthia Bir a,�, David Viano a, Scott Tashman b

a Bioengineering Center, Wayne State University, 818 W. Hancock, Detroit, MI 48202, USAb Department of Orthopaedic Surgery, University of Pittsburg, Pittsburg, PA, USA

a r t i c l e i n f o

Article history:

Accepted 24 July 2008The purpose of this study was to determine the force–time and force–displacement response of the

human mandible under direct loading at the chin. Sub-fracture response of the mandible and

Keywords:

Mandible

Impact biomechanics

Response corridors

90/$ - see front matter & 2008 Elsevier Ltd. A

016/j.jbiomech.2008.07.020

esponding author. Tel.: +1313 577 3830; fax:

ail address: cbir@wayne.edu (C. Bir).

a b s t r a c t

temporomandibular joint (TMJ) were analyzed from 10 cadavers that were impacted at the chin with a

2.8-kg mass at drop heights of 300, 400 and 500 mm and a 5.2-kg mass at 500 mm. Motion of radio-

opaque markers adhered to the surface of the bone was recorded at 1000 Hz by a bi-planar X-ray and

converted to three-dimensional coordinates. Peak force ranged from 0.90 to 4.54 kN causing chin

displacement of 1.2–4.4 mm. A bi-linear response was observed with stiffness of 475.17199.8 kN/m for

chin displacement resulting from loading up to 0.6 kN and 2381.67495.7 kN/m for loads from 0.6 to

3.25 kN. This defines the biomechanical response of the mandible for chin motion under impact loading.

The response of different segments of the mandible and TMJ are also documented. Force–time and

force–displacement response corridors for the mandible can be used for finite element model and/or

the development and validation of a biomechanical surrogate.

& 2008 Elsevier Ltd. All rights reserved.

1. Introduction

In efforts to document the conditions leading to mild traumaticbrain injury (MTBI) in American football, Pellman et al. (2003a, b)re-enacted 31 National Football League (NFL) cases with anthro-pometric test devices (ATD) to describe the head impact responseplayers were experiencing on the field. Four average impactconditions were described. Only one condition, Condition A,involved the potential for indirect loading of the mandiblethrough the chin strap.

Prior research has studied mandible fracture tolerance (Hodgson,1967; Nahum et al., 1968; Schneider and Nahum, 1972; Nahum,1975; Huelke and Compton, 1983; Hopper et al., 1994; Unnewehret al., 2003) and the response of the condylar fibrocartilage andtemporomandibular joint (TMJ) articulating disk to load (Lai et al.,1998; Hu et al., 2003; Beek et al., 2001; Tanaka et al., 2001).However, none of these studies described the gross deformation ofthe mandible versus time under chin loading. This response iscritical for developing a human surrogate headform with anarticulating jaw that can be used to evaluate football head andmouth protection equipment including mouthguards, facemasksand chin straps with regard to energy transferred via the jaw tothe skull that may lead to concussion.

ll rights reserved.

+1313 577 8333.

Response corridors have been widely used as a requirement forfinite element model and/or biomechanical surrogate validation(Kroell et al., 1971; Hardy et al., 2001; Stemper et al., 2001;Maltese et al., 2002; Bir et al., 2004; Wheeldon et al., 2006).The aim of this research was to develop force–time andforce–displacement response corridors of the mandible and TMJunder chin loading.

Others have studied biofidelity improvements to the Hybrid IIIATD headform (Newman and Gallup, 1984; Melvin and Shee,1989; Melvin et al., 1995) with emphasis on facial loadingresponse. This study provides the response requirements forloading at the chin that can be used for further refinement of theHybrid III or other surrogates.

2. Methodology

2.1. Chin impact test method development

The goal of this study was to document the response of the mandible to drop

mass impacts that produce comparable jaw loading to that seen in NFL Condition A

(Pellman et al., 2003a, b). A Hybrid III 50th percentile male headform was retrofit

by Robert A. Denton, Inc. (Rochester Hills, Michigan) for Biokinetics and Associates,

Ltd. (Ottawa, Canada) with a fixed triaxial load cell (Model 6080, Robert A. Denton,

Inc.) capable of measuring jaw loads (Fig. 1a). This headform was impacted in

Condition A using the helmeted surrogate pendulum impact methods described by

Pellman et al. (2006). The resulting force–time history data were then used to

establish a drop mass impact condition to the chin of the same Hybrid III headform

using a refined version of a test stand previously developed by Walilko (2004). This

drop mass condition would then be used on postmortem human subjects (PMHS)

ARTICLE IN PRESS

Electromagnet

75 mm Drop

Tube

Accelerometers

Nylon Cable

Drop Mass

Specimen

Foam Pad

80/20®

Aluminum Frame

Two 2.5 cm Thick Steel Plates

Head Fixture0.6 x 0.6 m, 1.3 cm Steel Table Top

4 Corner, 1 Center Support

z

x

Fig. 1. The 50th percentile Hybrid III with fixed triaxial load cell and test stand: (a) Hybrid III 50th male headform with triaxial jaw load cell (drawing courtesy of Robert A.

Denton, Inc.) and (b) Test stand.

M. Craig et al. / Journal of Biomechanics 41 (2008) 2972–2980 2973

ARTICLE IN PRESS

Table 1Specimens and usage

Specimen Age Sex Height (cm) Weight (kg) Specimen used in analysis Condition ]1

300 mm, 2.8 kg 400 mm, 2.8 kg 500 mm, 2.8 kg 500 mm, 5.2 kg

562 80 F 168 68 Noa�

c

31860 86 F 168 52 Noa�

c

31888 72 F 152 67 Nob�

31908 86 F 161 42 Noa� �

c

32009 78 F 178 68 Noa� �

c

31658 24 M 180 94 Yes � �c

31720 45 M 183 91 Yes �d

�

31923 55 M 183 122 Yes � � �

32042 85 M 152 79 Yes �

2204 71 M 180 75 Yes �

2228 57 M 175 88 Yes � �

2230 54 M 183 136 Yes � � � �c

2329 38 M 185 118 Yes � � � �

2334 70 M 180 54 Yes � � � �c

2345 42 M 165 82 Yes � � � �c

Total tests analyzed 5 9 6 6

a Specimen exhibited low fracture tolerance and/or joint dislocation at or below base 400 mm, 2.8 kg drop condition.b Specimen not included due to condition of dentition.c Fracture or joint subluxation occurred.d Tested at 350 mm/2.8 kg and 400 mm/5.2 kg, not 400 mm/2.8 kg.

Maxilla

High-speed X-ray

Acoustic Meatus

Fossa

Chin

Condyle

Temporal Bone

Body

Inferior view of mandible

Fig. 2. X-ray and image of specimen prepared with 2- and 3-mm radio-opaque markers.

Table 2Radio-opaque marker locations

Marker location No. of markers Diameter (mm)

Mental protuberance (chin) 1 3

LH/RH body/angle/ramus 12 2

LH/RH sub-condyle 2 3

Maxilla 2 2

External acoustic meatus 2 2

Superior to mandibular fossa 2 3

Temporal bone 8 3

M. Craig et al. / Journal of Biomechanics 41 (2008) 2972–29802974

for the purpose of collecting data to develop force–time and force–displacement

corridors.

Fig. 1b shows the test stand used in this study. The original stand had an

aluminum reaction surface. Steel plates and a center support were added to

reinforce the stand based on preliminary data.

2.2. PMHS testing

Table 1 summarizes the 15 PMHS heads that were used. The protocol was

approved by the Wayne State University Human Investigation Committee.

Specimens were treated in accordance with accepted ethical practices (King

et al., 1995). Specimens were required to have largely intact dentition given the

potential for reduced condylar bone strength with absent dentition (Giesen et al.,

2003). Specimens did not die from causes affecting the structures of the mandible,

TMJ or basilar skull.

Fig. 2 shows the images of a specimen prepared with radio-opaque markers

adhered to the surface of the bone using waterproof epoxy. A total of 29 markers

were used (Table 2). A high-speed, bi-planar X-ray system (Henry Ford Hospital,

Detroit, Michigan) recorded marker displacement at 1000 Hz using a pair of digital

cameras attached to image intensifiers. Two pairs of digital cameras were used

during the course of the study. The recorded image resolution was 1.7 pixel/mm

and 3.7 pixel/mm for the first and second pair of cameras, respectively. Proprietary

software (Henry Ford Hospital, Detroit, Michigan) was used to identify and track

markers for each set of two-dimensional images and to convert that data into a

three-dimensional representation of marker position over time. Marker displace-

ment synchronized with drop mass load was used to define the biomechanical

response of the mandible that forms the basis of the biofidelity requirements

necessary for surrogate validation. No fractures were observed at a marker

location, and it was judged that marker attachment did not affect the compliance

of the bony structures.

The PMHS heads were secured in a fixture with polyester resin (Bondos,

Bondo Corporation, Atlanta, Georgia). Two drop masses (2.8 and 5.2 kg) with a

20.3-cm2 circular impact surface were dropped onto the chin of the specimens.

Drop heights, orientation and impact attenuation were determined based on

testing with the jaw load cell Hybrid III headform. Heads were impacted multiple

times with drop mass and height combinations targeted at producing multiple

sub-fracture data points (Table 1). Drop mass acceleration was measured in the x-,

y- and z-directions with EndevcoTM 7264C-2000 single-axis translational accel-

erometers and collected using TDAS PROTM (DTS, Inc., Seal Beach, California) at

ARTICLE IN PRESS

Condition #1 : Midsagittal through chin and condyles

Condition #2 : Through chin and condyles, 30° lateral to Condition #1

z

x

z

y

Fig. 3. Drop mass impact orientations.

-2

-1

0

1

2

3

0.005

Forc

e (k

N) Fx - Condition A

Fx - Drop Stand

Fy - Condition AFy - Drop Stand

Fz - Condition AFz - Drop Stand

Force - Drop Mass

0.020.0150.01

M. Craig et al. / Journal of Biomechanics 41 (2008) 2972–2980 2975

10,000 Hz. Drop mass force was calculated as the product of the drop mass weight

and z-direction acceleration. Acceleration data were filtered using a 1650-Hz

Butterworth filter (SAE CFC 1000).

It was necessary for a specimen to be able to sustain the load of a baseline

impact without fracture or joint subluxation to qualify for inclusion in the

response corridors. The baseline impact would be the one that produced peak

forces, which, on average, were below the 1.89-kN fracture threshold seen by

Schneider and Nahum (1972).

To support multiple impacts, high-speed X-ray video, accelerometer data and

specimen analysis were completed between tests to confirm that no fractures, TMJ

joint subluxation or significant soft tissue damage had occurred before continuing

testing. As necessary, a repeat or baseline test was done to confirm that continued

testing was appropriate.

-3Time (s)

Fig. 4. Drop mass versus NFL Condition A pendulum.

3. Results

3.1. Chin impact test method development

The head orientation, drop mass weight and height that bestmatched NFL Condition A in impacts to the chin of the jaw load cellHybrid III headform was a 301 lateral impact with the plane of loadingthrough the condyles (Condition #2), 2.8 kg and 500 mm, respectively(Fig. 3). The matching test included an 18-mm thick, closed-cell, 3.8pcf polyethylene foam pad (Minicel& L380, Sekesui Voltek, LLC) on theimpacting surface of the drop mass. Fig. 4 shows the force on chinresponse of the matching drop mass test to a pendulum re-enactmentof Condition A. The loading duration compares well with the NFLMTBI cases of Pellman et al. (2003a) and the peak forces seen in theCondition A pendulum testing. The resultant jaw load cell forces were2.44 and 2.48 kN for the pendulum re-enactment and drop standvalidation tests, respectively.

Condition ]2 matched the peak responses and duration of NFLCondition A. However, it was determined that Condition ]1 wouldbe the impact condition used to produce response corridors(Fig. 3). Condition ]1 involved a mid-sagittal impact to the chinwith symmetric, stable loading directed through the condyles thatminimized mandible rotation and load transfer into the upperdentition and maxillae. Similar observations and methodsregarding the primary direction of force (PDOF) in chin impacttesting were made by Hodgson (1967), Nahum et al. (1968) andSchneider and Nahum (1972).

3.2. PMHS test results and corridor development

Twenty-six tests from 10 specimens were analyzed. Thebaseline or qualifying impact condition was determined to be a400-mm drop of the 2.8-kg mass, which saw average peak forcesof 16817442 N. Four female specimens failed to meet the baseline

condition due to fracture or joint subluxation and one dueto dentition. Reduced mechanical properties resulting fromadvanced age and/or osteoporosis may have contributed to thelow fracture tolerance. The remaining 10 specimens included inthe corridor development were all male.

Three specimens were tested on both the original and thereinforced stands with the 2.8-kg mass at 400 mm. On average,there was a 17% and 26% increase in peak force and chindisplacement, respectively, that was observed on the reinforcedversus original test stands. The percent difference over time wasused in part to scale data from specimens that were tested only onthe original stand (See Appendix A for details).

Displacement data were grouped into four main areas (Fig. 5):(1) chin, (2) average condyle (sagittal plane), (3) lateral deflectionof mandibular body, ramus and condyles, and (4) mandiblecompression between the chin and condyles. Measurements (1)and (2) were relative to a fixed marker on the temporal bone thatwas not subject to deformation under drop mass loading.

The intent of Condition ]1 was to establish a stable loading ofthe mandible. However, on average, 1.51 of jaw opening occurredabout the anatomical y-axis. The chin displacement was correctedto eliminate rotational displacement. A trend towards jaw openingas opposed to closing acted to minimize the unmeasured forcesthat would have been transferred between the upper and lowerdentition and was supported by an absence of tooth damage.

Table 3 shows the average peak force and displacement. Trendsinclude an increase in displacement with increased drop massheight and increasing lateral expansion of the mandible fromthe first marker on the body to the sixth marker on the ramus.The lateral movement at the condyles was comparable to thedisplacement at the ramus markers.

ARTICLE IN PRESS

Drop Mass(1) Chin Displacement

(2) Avg CondylarSagittal Disp.

Pre-test Marker Position

Marker Position at Peak Deformation

Chin

Condyle

(4) Chin to Condyle Compression

Six Body/Ramus Markers Placed Symmetrically on Either Side of Mandible

(3) Relative Lateral Displacement of Body/Ramus & Condyle Markers

Condyle Markers

Body/Ramus Markers

#1#2

#3#4

#5#6

Drop Mass

z

y

Fig. 5. X-ray image and diagram showing mandible displacement measurements.

Table 3Peak force and displacement for Condition #1

Condition Force (N) Displacement of chin and condyles

(mm)

Mandible body/ramus and condyle lateral expansion (mm)

Chin Average

condyle

(sagittal)

Chin to

condyle

Body Pt1 Body Pt2 Body Pt3 Body Pt4 BodyPt5

(gonion)

Body Pt6

(ramus)

Condyle

lateral

2.8 kg Mean 1242 1.95 1.62 0.28 0.15 0.28 0.27 0.38 0.54 0.56 0.71

300 mm Std dev. 253 0.60 0.51 0.11 0.09 0.06 0.09 0.05 0.16 0.15 0.44

2.8 kg Mean 1681 2.04 1.40 0.35 0.29 0.30 0.52 0.71 0.77 0.95 0.91

400 mm Std dev. 442 0.63 0.41 0.17 0.13 0.18 0.20 0.28 0.25 0.26 0.31

2.8 kg Mean 2293 2.53 2.02 0.40 0.30 0.44 0.60 0.77 1.04 1.20 1.05

500 mm Std dev. 256 0.70 0.60 0.18 0.10 0.19 0.20 0.30 0.40 0.57 0.55

5.2 kg Mean 3692 3.15 2.48 0.60 0.31 0.59 0.74 1.08 1.50 1.78 1.62

500 mm Std dev. 560 0.77 0.42 0.30 0.08 0.24 0.26 0.30 0.48 0.74 0.57

Note: Values reported up to peak force.

M. Craig et al. / Journal of Biomechanics 41 (2008) 2972–29802976

Fig. 6 shows force versus time for each Condition ]1 dropmass/height combination. Coordinates defining the force–timecorridors for each Condition ]1 combination are shown. Thesecoordinates follow at approximately 71.5 standard deviationsfrom the mean response.

Fig. 7 shows force versus chin displacement for the fourspecimens tested in all four drop mass and height combinations:2230, 2329, 2334 and 2345. No viscous behavior of the mandibleor related soft tissue and structures was evident in comparingimpact combinations. The data for each specimen are super-imposed within the limits of the experimental measurements. Theexception is the 500 mm, 5.2 kg test on 2329, which had a shorterinitial loading phase. This was likely due to a difference in theinitial positioning of the mandible. However, the slope of thesecond phase of loading at 2.2 kN/mm was comparable to that ofthe 2.8-kg tests at 2.570.2 kN/mm.

Peak forces ranged from 0.90 to 4.54 kN causing chindisplacement of 1.2–4.4 mm in the 26 analyzed tests. Theforce–displacement corridor shown in Fig. 7 is at approximately1.5 standard deviations from the mean response of all 26Condition ]1 tests. The data and corridor shown represents theresponse of the mandible up to peak drop mass load.

A bi-linear response was observed with an average stiffness of475.17199.8 kN/m for chin displacement with loads up to 0.6 kNand 2381.67495.7 kN/m for loads from 0.6 to 3.25 kN. Theaverage deflection of the chin at 600 N was 1.570.5 mm asderived through linear interpolation of the data. Drop mass

energy early in the event was primarily absorbed by the soft tissueof the chin and TMJ. From 600 to 3250 N there was only1.470.3 mm additional deformation of the chin as proportionallymore energy was absorbed by the stiffer structures of themandible and basilar skull.

Combined foam pad and test stand stiffness was evaluated indrops directly onto the test stand. The stiffness calculated fromdrop mass displacement up to peak load for a series of 500 mm,2.8 kg impacts was 1228.1798.1 kN/m. The mandible/skull stiff-ness derived from chin displacement and drop mass force in500 mm, 2.8 kg impacts was 690.77242.3 kN/m. The predictedeffective stiffness of the overall system in 500 mm, 2.8 kg PMHSimpacts would be approximately 442.1 kN/m, assuming negligiblestiffness of the soft tissues covering the chin. This value compareswell with the effective stiffness calculated from drop massdisplacement and force as measured in the 500, 2.8 kg PMHStests at 485.57106.9 kN/m.

Fig. 8 shows that the mandible loading has three distinctphases. Also shown are high speed X-ray images of the condylerelative to the mandibular fossa at the beginning and end of eachphase. The first phase shows similar chin and condyle motion withminimal compression of the mandible. In the second phase, chinto condyle mandible compression is present as condylar move-ment toward the mandibular fossa is met with greater resistance.Proportionately more energy is being absorbed by the bonystructures of the mandible and skull as TMJ joint space is reduced.The relative difference between chin and condyle motion is

ARTICLE IN PRESS

0.018, 220

0.014, 650

0.003, 375

0.007, 700

0.011, 1600

0.017, 100

0.014, 3600.007, 3200.003, 50

0.011, 850

0

1

2

0 0.005 0.01 0.015 0.02

Time (s)

Forc

e (k

N)

0.017, 300

0.013, 10500.007, 1000

0.01, 2200 0.011, 2200

0.005, 260 0.014, 180

0.01, 900

0

1

2

3

0 0.005 0.01 0.015 0.02

Time (s)

Forc

e (k

N)

0.017, 160

0.013, 8000.006, 800

0.009, 2650 0.01, 2650

0.01, 1750

0.015, 1000.012, 500

0.008, 825

0.003, 800

1

2

3

0 0.005 0.01 0.015 0.02Time (s)

Forc

e (k

N)

2204 2228 2230 2329 2334 234531658 31720 31923 32042 Mean Corridor

0.009, 5250

0.006, 1550

0.007, 300

0.009, 1300

0

1

2

3

4

5

6

0 0.005 0.01 0.015 0.02

Time (s)

Forc

e (k

N)

Mandible Fracture at 9 ms in 4 of 6 Specimens

Fig. 6. Condition ]1 force versus time data and corridors: (a) 300 mm, 2.8 kg, (b) 400 mm, 2.8 kg, (c) 500 mm, 2.8 kg and (d) 500 mm, 5.2 kg.

0

1

2

3

4

0 1 2 3 4Displacement (mm)

Forc

e (k

N)

2329 2230 2334 2345 CorridorMean – All Specimens

300 mm, 2.8 kg400 mm, 2.8 kg500 mm, 2.8 kg500 mm, 5.2 kg

3.9, 32500.7, 600

2.2, 600

1.5, 3250

Fig. 7. Condition ]1 force versus chin displacement data for specimens 2230, 2329,

2334 and 2345 and average response and corridor for all specimens.

M. Craig et al. / Journal of Biomechanics 41 (2008) 2972–2980 2977

equivalent to the amount of mandible compression that isoccurring between the chin and condyles. The final phaserepresents unloading of the mandible where mandible compres-sion returns to 0 mm and chin deformation is equal to residualcondyle deformation.

Eight specimens experienced fractures as a result of Condition]1 testing. Seven specimens experienced unilateral fracturesthrough the condylar neck and/or head. One saw bilateralfractures of the condylar heads. One specimen experienced afracture of the body of the mandible near the left mental tuberclethat occurred after an angulated fracture of the right condylarneck. Two specimens experienced fractures of the temporal boneat the anterior ear canal and a third at the posterior ear canal.A single specimen, which also saw condylar fracture, was found tohave unilateral joint subluxation or dislocation upon completionof the autopsy. The four female specimens (82.574.1 years of age)had an average fracture force of 1.6470.1 kN. The four malespecimens (48.3719.8 years of age) had an average fracture forceof 3.8370.7 kN. The average fracture load for females in thecurrent study was significantly less than for males (t ¼ 5.92,p ¼ 0.004). However, the age of those female specimens wassignificantly higher (t ¼ 3.38, p ¼ 0.021) than the male specimens.Fracture or subluxation occurrence and timing were evident bylocalized peaks in the force data that corresponded to visible andmeasurable changes in displacement of the markers on themandible.

4. Discussion and conclusions

Cadaveric testing is accepted as a valuable methodologyfor collecting biomechanical data to be used in establishing

ARTICLE IN PRESS

0 ms 3 ms

Condyle

Fossa

20 ms 9 ms

3 mm CondyleMarker

High-speed X-ray

TMJ 0.0

0.5

1.0

1.5

2.0

0.0200.0100.0050.000Time (s)

Dis

plac

emen

t (m

m)

0

0.5

1.0

1.5

2.0

2.5

Forc

e (k

N)

Force

Mandible Compression

0.015

TMJ Joint Space

Phase 1

ChinCondyle

Inferior / Oblique Phase 3Phase 2

Fig. 8. Force and displacement versus time and high-speed X-ray images for the 500 mm, 2.8 kg Condition ]1 impact on specimen 2329.

M. Craig et al. / Journal of Biomechanics 41 (2008) 2972–29802978

biofidelity requirements, but has recognized limitations (Kinget al., 1995). Specific concerns include the age of the generallyolder PMHS population and changes in the compliance of boneand soft tissue. The specimen acceptance criteria and qualifyingimpact condition helped address this limitation by ensuring aminimum strength for all specimens included in the responsecorridors. There is not an acceptable substitute for PMHS indocumenting the response of the human mandible under chinloading due to the complexity of the mandible and TMJ.

The average age of the 10 specimens used to develop theresponse corridors was 54718 years with seven under 60 years ofage. Fig. 9a shows the distribution of age and fracture/dislocationload for the current study and results of simple linear regressionshowing a significant negative correlation between age and forcerequired to produce fractures. However, there was no significantcorrelation found between age and drop mass force (R ¼ 0.038,t ¼ �0.100, p ¼ 0.462), chin displacement (R ¼ 0.255, t ¼ �0.698,p ¼ 0.254), gonion lateral movement (R ¼ 0.476, t ¼ 1.433,p ¼ 0.097), chin to condyle mandible compression (R ¼ 0.115,t ¼ 0.285, p ¼ 0.393), or condyle sagittal displacement (R ¼ 0.210,t ¼ �0.568, p ¼ 0.294) for the sub-fracture response data.

Mandible fracture tolerance in the current study can becompared to the data from Hodgson (1967) and Schneider andNahum (1972). Fig. 9b shows the distribution of male specimenswith fractures for current and prior studies (Hodgson, 1967;Schneider and Nahum, 1972). The current study had significantlyhigher average fracture force (t ¼ 3.30, p ¼ 0.011) for malespecimens versus those of Hodgson (1967) that fractured at2.3670.5 kN. The significant difference may be attributed to thehigher age (t ¼ 2.36, p ¼ 0.032) of Hodgson’s specimens versusthe current study, 76.0712.6 versus 48.3719.8 years of age,respectively. Comparing male specimens in the current andSchneider and Nahum (1972) studies showed no statisticallysignificant difference in mean fracture force (t ¼ 1.19, p ¼ 0.144)at 3.8370.7 versus 3.0671.1 kN, respectively.

Hopper et al. (1994) dropped cadaver heads chin first ontofixed surfaces and saw average fracture forces of 5.2770.93 kN.They oriented the head such that the PDOF was through the headcenter of gravity (CG). The angle between the chin and the headCG in the sagittal plane is approximately 541 from horizontalbased on skull landmark coordinate data from Hubbard andMcLeod (1973) that were derived from measurements taken byByars et al. (1970). The angle from the chin to the condylionsderived from the current analysis of those studies was approxi-mately 471 from the horizontal. In the current study, the samemeasure averaged approximately 491. The greater vertical anglebetween the chin and the head CG versus the chin and thecondyles results in greater energy transfer between the upper andlower dentition. This would contribute to a higher force beingrequired to fracture the mandible.

Multiple tests per specimen were necessary to collect asignificant amount of sub-fracture data. While micro-cracks maybe a concern in repeated loading of bony structures, it is notbelieved to have influenced this study. As noted, the averagefracture loads were comparable to those of Schneider and Nahum(1972) and significantly greater than those of Hodgson (1967).Additionally, the average loads seen in the 300, 400 and 500 mm,2.8 kg impacts at 1.1970.2, 1.6570.5 and 2.2570.3 kN, respec-tively, were below the average fracture loads for male specimensof 3.8770.6 and 3.0671.1 kN documented in this and theSchneider and Nahum studies, respectively. Finally, the protocolfor continued testing involving evaluation of the specimen, X-rayvideo and accelerometer data between tests and completion ofrepeat or baseline tests assured that no fractures or othersignificant damage had occurred. Given the comparable force–displacement trends seen in repeat testing of specimens (Fig. 7)and the comparable or greater average fracture loads versus priorstudies, it appears that the cumulative effects of repeat testingdid not alter the response of the mandible and surroundingstructures.

ARTICLE IN PRESS

F = -48.93x + 5930.8R2 = 0.7638, t = -4.405, p = 0.003

0

1

2

3

4

5

Age (years)

Forc

e (k

N)

MaleFemale

F = -35.734x + 5405.1R2 = 0.4415, t = -2.811 , p = 0.009

0

1

2

3

4

5

Age (years)

Forc

e (k

N)

Current StudySchneider and Nahum (1972) Hodgson (1967)

9080706050403020100

9080706050403020100

Fig. 9. Fracture force versus specimen age: (a) Specimens with fracture/dislocation-current and (b) Male specimens: current and prior studies.

M. Craig et al. / Journal of Biomechanics 41 (2008) 2972–2980 2979

4.1. Sources of variation

Anatomical differences between PMHS contributed to variationin measurements in the current study. For example, Preti et al.(1982) found the standard deviation in distance of the centercondyle to the mandibular fossa to be approximately 0.8 mm atmaximum intercuspation. Pre-test condylar distance to themandibular fossa was not measured in the current research.However, it is expected that differences in initial condylar positionwould have contributed to variation in peak chin and condylardisplacement.

Variation also existed in determining the kinematic data frommarker tracking. The standard deviation of the distance betweentwo markers with a fixed distance between them was measuredover time in analyzed cases. The average standard deviation orprecision of the measurement between the fixed points was0.05470.03 mm, which is comparable to the precision of0.064 mm documented by Tashman and Anderst (2003) usingthe same system. Additionally, Tashman and Anderst (2003)found no bias in the measurements. Thus, the accuracy of themeasurements of the current study would be equivalent to theprecision or 0.054 mm. Maxilla, external acoustic meatus andmandibular fossa marker displacements were not reported,because tracking error represented a large percentage of less than0.2-mm peak displacements measured in those locations.

4.2. Conclusions

The force–time and force–displacement corridors documentedin Figs. 6 and 7 together define the response of the human

mandible to direct loading of the chin. The high-speed X-ray andtracking methods quantified the gross deformation of differentsegments of the mandible and TMJ under loading. These resultsfrom Condition ]1 testing show how the mandible and TMJdeform under stable, symmetric loading. While the NFL ConditionA impact involves oblique loading of the mandible, the loadingduration for PMHS Condition ]1 tests compares well. Also, theevaluated range of Condition ]1 drop mass loads overlaps the2.6 kN observed in validating the test methods against NFLCondition A (Fig. 4). The displacement of the chin, together withthe segments of deformation observed in the mandible and TMJ,provides a new view of the overall compliance or shock-absorbingcharacteristics of the mandible and TMJ.

The force–time and force–displacement corridors can serve asthe biomechanical response requirements for a human surrogateand/or finite element model. This study documented stiffness ofthe human mandible in testing where interaction between theupper and lower dentition was minimized. Impact orientation andinteraction between the upper and lower dentition can contributeto differences in the responses of the mandible. As such, thecorridors presented here may be used in the validation of a humansurrogate with an articulating jaw, but additional testing thatincludes greater energy transfer between the upper and lowerdentition may be warranted.

Conflict of interest statement

None of the authors has a financial interest or businessrelationship posing a conflict of interest concerning the researchon the mandible response to impact loading.

ARTICLE IN PRESS

M. Craig et al. / Journal of Biomechanics 41 (2008) 2972–29802980

Acknowledgements

The authors gratefully acknowledge the staff at the Henry FordHospital Bone and Joint Center, Motion Analysis Laboratory fortheir assistance with test setup and image processing, Biokineticsand Associates, Ltd., for completing surrogate impactor tests usedin the validation of test methods, and the staff at the Wayne StateUniversity Bioengineering Center for their assistance in complet-ing testing. This research was supported by NFL Charities.

Appendix A. . Supplementary materials

The online version of this article contains additional supple-mentary data. Please visit doi:10.1016/j.jbiomech.2008.07.020

References

Beek, M., Aarnts, M.P., Koolstra, J.H., Feilzer, A.J., van Eijden, T.M., 2001. Dynamicproperties of the human temporomandibular joint disc. Journal of DentalResearch 80 (3), 876–880.

Bir, C.A., Viano, D.C., King, A.I., 2004. Development of biomechanical responsecorridors of the thorax to blunt ballistic impacts. Journal of Biomechanics 37(1), 73–79.

Byars, E.F., Haynes, D., Durham, T., Lilly, H., 1970. Craniometric measurements of humanskulls. In: 1970 ASME Winter Annual Meeting. ASME Paper no. 70-WA/BHF-8.

Giesen, E.B., Ding, M., Dalstra, M., van Eijden, T.M., 2003. Reduced mechanical loaddecreases the density, stiffness, and strength of cancellous bone of themandibular condyle. Clinical Biomechanics 18 (4), 358–363.

Hardy, W.N., Schneider, L.W., Rouhana, S.W., 2001. Abdominal impact response torigid-bar, seatbelt, and airbag loading. Stapp Car Crash Journal 45, 1–31.

Hodgson, V.R., 1967. Tolerance of the facial bones to impact. American Journal ofAnatomy 120, 113–122.

Hopper, R.H., McElhaney, J.H., Myers, B.S., 1994. Mandibular and basilar skullfracture tolerance. In: Proceedings of the 38th Stapp Car Crash Conference, SAEPaper No. 942213, Ft. Lauderdale, FL.

Hu, K., Qiquo, R., Fang, J., Mao, J.J., 2003. Effects of condylar fibrocartilage on thebiomechanical loading of the human temporomandibular joint in a three-dimensional, nonlinear finite element model. Medical Engineering & Physics25 (2), 107–113.

Hubbard, R.P., McLeod, D.G., 1973. A basis for crash dummy skull and headgeometry. In: King, W.F., Mertz, H.J. (Eds.), Human Impact Response—

Measurement and Simulation. Plenum Press, New York, pp. 129–152.Huelke, D.F., Compton, C.P., 1983. Facial injuries in automobile crashes. Journal of

Oral and Maxillofacial Surgery 41 (4), 241–244.King, A.I., Viano, D.C., Mizeres, N., States, J.D., 1995. Humanitarian benefits of

cadaver research on injury prevention. The Journal of Trauma-Injury Infection& Critical Care 38 (4), 564–569.

Kroell, C.K., Schneider, D.C., Nahum, A.M., 1971. Impact tolerance and response ofthe human thorax. In: Proceedings of the 15th Stapp Car Crash Conference, SAEPaper No. 710851, San Diego, CA.

Lai, W.F., Bowley, J., Burch, J.G., 1998. Evaluations of shear stress of the humantemporomandibular joint disc. Journal of Orofacial Pain 12 (2), 153–159.

Maltese, M.R., Eppinger, R.H., Rhule, H.H., Donnelly, B.R., Pintar, F.A., Yoganandan,N., 2002. Response corridors of human surrogates in lateral impacts. Stapp CarCrash Journal 46, 321–351.

Melvin, J., Shee, T., Facial injury assessment techniques. In: Backaitis, S.H., Mertz,H.J. (Eds.), Hybrid III: The First Human-like Crash Test Dummy. SAE Paper No.896072, Warrendale, PA, 1989, pp. 453–462.

Melvin, J.W., Little, W.C., Smrcka, J., Zhu, Y., Salloum, M.J., 1995. A biomechanicalface for the Hybrid III dummy. In: Proceedings of the 39th Stapp Car CrashConference, SAE Paper No. 952715, San Diego, CA.

Nahum, A.M., 1975. The biomechanics of facial bone fracture. Laryngoscope 85 (1),140–156.

Nahum, A.M., Gatts, J.D., Gadd, C.W., Danforth, J., 1968. Impact tolerance of theskull and face. In: Proceedings of the 12th Stapp Car Crash Conference, SAEPaper No. 680785, Detroit, MI.

Newman, J.A., Gallup, B.M., 1984. Biofidelity improvements to the Hybrid IIIheadform. In: Proceeding of the 28th Stapp Car Crash Conference, SAE PaperNo. 841659, Chicago, IL.

Pellman, E.J., Viano, D.C., Tucker, A.M., Casson, I.R., Waeckerle, J.F., 2003a.Concussion in professional football: reconstruction of game impacts andinjuries. Neurosurgery 53 (4), 799–814.

Pellman, E.J., Viano, D.C., Tucker, A.M., Casson, I.R., 2003b. Concussion inprofessional football: location and direction of helmet impacts–part 2.Neurosurgery 53 (4), 1328–1340.

Pellman, E.J., Viano, D.C., Withnall, C., Schewchenko, N., Bir, C.A., Halstead, P.D.,2006. Concussion in professional football: helmet testing to assess impactperformance—part 11. Neurosurgery 58 (1), 78–96.

Preti, G., Scotti, R., Turbiglio, G., Scala, C., 1982. Anatomic study of condylar positionat maximum intercuspation. Journal of Prosthetic Dentistry 47 (4), 445–448.

Schneider, D.C., Nahum, A.M., 1972. Impact studies of facial bones and skull. In:Proceedings of the 16th Stapp Car Crash Conference. SAE Paper No. 720965,Detroit, MI.

Stemper, B.D., Yoganandan, N., Pintar, F.A., Sun, Z., 2001. Development of extensionkinematic corridors to validate a head/neck finite element model. BiomedicalSciences Instrumentation 237, 239–244.

Tanaka, E., Sasaki, A., Tahmina, K., Yamaguchi, K., Mori, Y., Tanne, K., 2001.Mechanical properties of human articular disk and its influence on TMJ loadingstudied with the finite element method. Journal of Oral Rehabilitation 28 (3),273–279.

Tashman, S., Anderst, W., 2003. In-vivo measurement of dynamic joint motionusing high speed biplane radiography and CT: application to canine ACLdeficiency. Journal of Biomechanical Engineering 125 (6), 238–245.

Unnewehr, M., Homann, C., Schmidt, P.F., Sotony, P., Fischer, G., Brinkmann, B.,Bajanowski, T., DuChesne, A., 2003. Fracture properties of the humanmandible. International Journal of Legal Medicine 117 (6), 326–330.

Walilko, T.J., 2004. Biomechanical response of the temporomandibular joint fromimpacts in boxing, PhD. Thesis, Wayne State University, Detroit, MI.

Wheeldon, J.A., Pintar, F.A., Knowles, S., Yoganandan, N., 2006. Experimentalflexion/extension data corridors for validation of finite element modelsof the young, normal cervical spine. Journal of Biomechanics 39 (2),375–380.