DEVELOPMENT AND VALIDATION OF A FINITE ELEMENT MODEL OF THE
SUPERIOR GLENOID LABRUM
Christopher J Gatti, Joseph D Maratt, Mark L Palmer, Richard E Hughes, James E Carpenter
University of Michigan, Ann Arbor, MI, USA
email: firstname.lastname@example.org, web: www-personal.umich.edu/~rehughes/index.html
The glenoid labrum is a common source of
musculoskeletal pain and disability. Superior
humeral head translation is one proposed
mechanism of injury. The labrum is difficult to
study in a laboratory setting because of its small
size and articular location. The purpose of this study
was to develop and validate a finite element model
of the glenoid labrum for humeral head translation
in order to understand the mechanical environment
experienced by the labrum.
The geometry of the glenoid, glenoid cartilage,
labrum, and humeral head cartilage was obtained
from μCT imaging (voxel size of 93 μm,
reconstructed at 186 μm) and segmented using
Mimics (Materialise NV). Boolean operations were
used to distinguish the labrum from the glenoid
cartilage and obtain accurate geometry of each
tissue. Tetrahedral meshes were created in
HyperMesh 9.0 (Altair Engineering, Inc.) using
C3D4 elements for the glenoid labrum and cartilage
(26,177 and 23,141 elements, respectively) and
R3D3 rigid elements for the glenoid bone and
humeral head cartilage (3,772 and 7,729 elements
respectively) (Fig. 1).
Figure 1: Finite element meshes of the labrum,
glenoid cartilage, and humeral cartilage.
The labrum was modeled using a homogeneous,
linear elastic, transversely isotropic material to
represent the circumferentially-oriented fibrous
composition of the labrum (Ep=0.24 MPa ,
Eθ=22.8 MPa , νp=0.33, νθp=0.10, Gθp=2.0 MPa
, where p defines the transverse plane and θ
defines the circumferential direction of the labrum).
The transversely isotropic material law was
implemented using coordinate systems aligned to
the circumferential orientation of the labrum. The
glenoid cartilage was modeled as a homogeneous,
linear elastic, isotropic material (E=1.7 MPa ,
ν=0.018 ), and the glenoid bone and humeral
cartilage were modeled as rigid surfaces. The rigid
glenoid surface was fixed in all degrees of freedom.
The humeral head cartilage was oriented such that
the humerus was positioned in 30° of glenohumeral
abduction in the scapular plane with zero humeral
rotation. The humeral cartilage was constrained
from rotation and loaded with a 50N compression
force directed into the glenoid. Quasi-static analyses
were performed by translating the humeral head
superiorly along the glenoid using displacement-
control for 1, 2, and 3 mm above the center of the
ABAQUS/Explicit v6.9 (SIMULIA, Inc.) with a
critical time step of approximately 5e-6 seconds.
A validation experiment was performed on 6
shoulder specimens in a custom testing fixture. The
boundary and loading conditions were set to be
easily reproduced in the finite element model. The
glenoid was fixed with the articular surface directed
upward. A 50N compressive load was applied to the
humerus directed into the glenoid, and the humerus
was translated superiorly along the glenoid using
computer-controlled motors. Radiopaque beads
were placed along the superior labrum to determine
the amount of movement of each bead in the plane
of the glenoid. Radiographs were taken at 1, 2, and
3 mm of superior humeral head translation. The
glenoid-plane movement of nodes along the
superior labrum in the finite element model were
then compared to the glenoid-plane movement of
the beads in the validation experiment.
were run using
RESULTS AND DISCUSSION Download full-text
The results of the model compared well to those
from the validation experiment, especially in the
superior and posterosuperior regions of the labrum
(Fig. 2). The strains were greatest in 11-12 and 12-1
regions of the labrum and generally increased with
greater humeral head translations (Fig. 3, 4).
Labrum displacement (mm)
Figure 2: Labrum displacement for the finite
element model and validation experiment for 3 mm
of superior humeral head translation.
Figure 3: Strains in the labrum for 0, 1, 2, and 3
mm of superior humeral head translation (A, B, C,
The labrum and biceps attachment have high
morphological variability which made standardizing
a testing protocol between the model and validation
experiment challenging . Although the finite
element model was based on a single specimen, the
results of the validation experiment based on 6
specimens compared well to the model without
optimizing model parameters. The displacement
measures of the labrum were compared for
movement only in the plane of the glenoid, and it is
possible that there was movement out of this plane
which was not measured due to the single
radiograph technique. The model could be
improved by using hexahedral elements and
refining the integration of the biceps tendon, and
this would likely extend the usability of this model.
9−10 10−11 11−1212−11−2 2−3
Figure 4: Mean (± 1 st. dev.) strains the superior
labrum for 0, 1, 2, and 3 mm of superior humeral
A finite element model of the glenoid labrum was
developed and validated for humeral head
translation. This model could be used to better
understand the local mechanical environment of the
labrum and its relationship to pathology.
 Carey J, et al. J Biomed Mat Res 51, 711-716,
 Smith CD, et al. J Anat 212, 49-54, 2008.
 Yao J, et al. J Biomech Eng 128, 135-141, 2006.
 Cohen B, et al. Transactions of ORS, 1993.
 Smith C, et al. Shoulder & Elbow doi:
This work was funded by the Department of
Orthopaedic Surgery and the Valassis Endowed
Research Fund. The authors thank all those who
contributed to this study.