A SYNERGISTIC APPROACH TO TRANSTIBIAL SOCKET INTERFACE MECHANICS:
EXPERIMENTS AND MODELING
By
Amy Lorraine Lenz

A DISSERTATION
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
Engineering Mechanics – Doctor of Philosophy
2017

ABSTRACT
A SYNERGISTIC APPROACH TO TRANSTIBIAL SOCKET INTERFACE MECHANICS:
EXPERIMENTS AND MODELING
By
Amy Lorraine Lenz
Prosthetics is a clinical field in need of further investigation for the improvement of patient
care. Engineering principles can be used in collaboration with clinical expertise to quantify key
mechanical issues occurring at the residual limb to prosthetic socket interface. Deep penetrating
ulcers can form on the residual limb within the socket and the formation is not understood in the
current research regarding interface mechanics. Quantitative data on limb motion within the
socket, shear forces at the interface, and propagation of these loads to the skin level and deeper
tissues are all lacking in current literature. The broad goal of this research was to understand the
interface mechanics of the gel liner on the residual limb relative to the prosthetic socket to improve
our understanding of displacements, loads and gel liner slip or no slip conditions.
This work consisted of four aims. Objective 1: Develop a quantitative method for assessing
motions between the prosthetic device and gel liner on the residual limb for patients with transtibial
amputation. Objective 2: Determine limb displacements, strains, relative socket to limb
displacements and angular rotations of transtibial limbs within a prosthetic socket during gait.
Objective 3: Quantify normal and shear force within the prosthetic socket for use in modeling.
Objective 4: Determine the level of tissue stresses within a layered finite element model including
gel liner interactions, constrained with experimental conditions of displacement and normal force.
First, a method to obtain kinematics within a socket was developed using motion capture
thin-disc markers beneath the surface of a clear prosthetic socket. Results comparing motion
capture with gold standard measurements statistically supported the use of this method.

Secondly, the newly developed method was used to obtain limb displacements, strains,
relative socket to limb displacements and angular rotations within a prosthetic socket during gait
from eight participants. Reflective markers with motion capture were used to track displacements
of the gel liner located within the clear prosthetic socket device. Results provide the most
comprehensive data set of interface kinematics in a transtibial amputee population and
significantly contribute to knowledge of interface mechanics which are a direct predictor of ulcer
formation.
Thirdly, a single transtibial prosthetic socket was instrumented with a two axis load cell to
measure kinetics at the internal socket wall. The participant walked in three conditions: gel liner,
three ply sock and a hole cut through the liner to measure forces at the skin. Shear and normal
force data were obtained during walking for these three conditions.
Lastly, simulations of tissue layers in transtibial amputees were modeled with Finite
Element Methods in FEBio. The gel liner to skin interface was modeled for two situations 1) gel
liner slips on the skin or 2) does not slip relative to the skin. Kinematic and kinetic conditions
obtained in earlier objective served as boundary conditions.

The purpose was to further

understand tissue stresses that may lead to pressure ulcer development and evaluate the
influence of various liner stiffness and thicknesses on underlying tissue stresses.
The presented research benefits the biomechanical community by addressing multiple
gaps in the literature and our understanding of the interface mechanics associated with
prosthetics. These data also further our understanding of how pressure ulcer formation may
progress due to internal resulting stresses.

Copyright by
AMY LORRAINE LENZ
2017

This dissertation is dedicated to the pursuit of worldwide dreams.

v

ACKNOWLEDGEMENTS

As a wise mentor once told me, the path to a PhD is a sinusoidal road with a fair share of
hills and valleys. As I reflect on those key moments, I am grateful and blessed to have such
wonderfully supportive people in my life who have each contributed to my success on this difficult
path.
This PhD research would not have been possible without the guidance of my advisor,
Tamara Reid Bush. Thank you for believing in me and giving me the opportunity to work on such
a novel and stimulating project. Combining your background in pressure ulcers and mine in
human gait mechanics lead to a fun adventure in prosthetics for both of us. I greatly appreciate
all you have taught me over the years, especially the important skill of saying no. It has been
very rewarding learning from you. I will always be grateful for your willingness to work with me
as a non-traditional PhD student and be fond of the work we completed during this time.
I would like to thank my wonderful committee members Dr. Roccabianca, Dr. Kwon and
Dr. Bix for their support throughout my dissertation development. It has been a pleasure learning
from them and receiving their feedback to improve my dissertation. Thank you for their generous
time given to support my research.
A huge thank you is owed to all of my lab members (Wu Pan, Amber Cussen, Josh Drost,
Jessica Buschman and Justin Scott) for supporting me during my soap opera life throughout these
past years and providing a helpful ear with meaningful advice to successfully make it through.
Funding support made the success of this work possible, especially the gracious
fellowship contributions from Michigan State University Graduate Office and Michigan State
University Mechanical Engineering Department. I am extremely thankful for this support which
helped me through challenging unexpected downturns financially.
vi

The Society of Women

Engineers have always been a source of encouragement and their scholarship support
contributed to my success as well. Thank you also to the School of Engineering at Grand Valley
State University for their support in my professional development through equipment and
resources. Lastly, my dissertation could not have been possible without the collaboration at Mary
Free Bed Rehabilitation Hospital and clinical collaborators, especially Katie Johnson and Patrick
Logan.

All the staff and management in the Orthotics and Prosthetics division have been

supportive in supplying resources for my experimental designs including thermoplastics, scrap
materials and duplication of prosthetic sockets.
To my loving dog, Crystal: Here’s to the long nights of you watching me work at my desk
(both at home and in the lab). Your continuous snuggles, kind kisses and gazing eyes of belief
have gotten me through years of school. Your never ending support is worth far more than the
small adoption fee I once paid, because you’re part of me in life and spirit. We’ve moved
mountains together and gotten through many moves across valleys. The support of a dog is
unparalleled when it comes to selflessness. “A dog is the only thing on earth that loves you more
than she loves herself.”
To my mom, Karen, and dad, Russell: I cannot thank you enough for all the support and
encouragement you’ve given me over the years. Mom, you have been my partner in crime, my
steady rock, my personal cheerleader and my tireless source of inspiration. Thank you for always
coming to my endless rescue at the drop of a dime. We’ve had so many priceless memories that
have gotten me through the ten years of school. Here’s to more of them and forever being
appreciative of your contributions to my career success. Dad, you have always been my tinkering
inspiration that began my journey as an engineer at a very young age. Thank you for instilling
such great qualities in me which have allowed me to grow as a successful engineer. Even when
you didn’t realize, you were teaching me how to think outside the box, critically evaluating
problems and developing creative solutions.

This dissertation was developed on a clinical
vii

curiosity that I believed engineering could help to solve. Thank you dad for having encouraged
me at a young age to think in this manner. It will forever be useful and essential to my career in
research and engineering.
To my irreplaceable friend, John: Thank you for being such as support as an expert in
SolidWorks, stable tower of strength and late night sounding board. You’ve been essential to the
success of my research, passing qualifying exams and motivating homework for the entire length
of this degree and treacherous road. I wouldn’t have been able to do it without you.
To my amazing engineering interns, Alex, Dayna, Alyssa, Kate, Sydney, Curtis, Alyna and
Brooke: It was extremely rewarding mentoring you and working with you over the years. I’m so
proud to see how you’ve grown, developed professionally and are all succeeding so well in your
engineering careers. You each contributed to my dissertation research by stimulating valuable
technical questions, being an extra set of hands for data collection, processing in Vicon, having
fun in MATLAB and troubleshooting in FEBio. I’m thankful to have had the opportunity to work
with all of you.

viii

TABLE OF CONTENTS

LIST OF TABLES .................................................................................................................................... xii
LIST OF FIGURES ..................................................................................................................................xiv
KEY TO ABBREVIATIONS................................................................................................................... xvii
1.

INTRODUCTION ................................................................................................................................ 1
1.1 Overview .......................................................................................................................................... 2
1.2 Limb Loss Statistics ..................................................................................................................... 5
1.3 Skin Disorders in Amputees....................................................................................................... 5
1.4 Pressure Ulcer Statistics ............................................................................................................. 6
1.5 Peripheral Arterial Disease: Risk Factors for Foot Ulceration and Implications in
Diabetic Patients .................................................................................................................................. 7
1.5.1 Importance of Restructuring Blood Flow After Amputation ................................................ 7
1.5.2 Diabetic Foot ............................................................................................................................. 8
1.5.3 Reasons for Amputation .......................................................................................................... 9
1.5.4 Angiosomes............................................................................................................................... 9
1.6 Prosthetic Device Components: Fabrication Selections .................................................. 10
1.6.1 Sockets .................................................................................................................................... 11
1.6.2 Socket Interface ...................................................................................................................... 12
1.6.3 Suspension .............................................................................................................................. 12
1.6.4 Combinations of Socket Interface and Suspension .......................................................... 14
1.7 Experimental Literature on Lower Extremity Amputees ................................................... 15
1.7.1 Interface Pressures ................................................................................................................ 15
1.7.2 Shear Stress at Interface ...................................................................................................... 16
1.7.3 Friction and Prosthetic Liners ............................................................................................... 17
1.7.4 Pylon to Socket Interface: Load Cells, Forces, Moments, Inverse Dynamics .............. 18
1.7.5 Kinematics: Whole body and Limb within Socket .............................................................. 18
1.7.6 Residual Limb Volume Changes.......................................................................................... 20
1.8 Finite Element Modeling of Residual Limb and Deep Tissue Injury .............................. 21
1.9 Conclusions .................................................................................................................................. 22

2. A NEW METHOD TO QUANTIFY RESIDUAL LIMB MOTION WITHIN A PROSTHETIC
SOCKET FOR BELOW KNEE AMPUTEES ....................................................................................... 24
2.1 Abstract.......................................................................................................................................... 25
2.2 Introduction................................................................................................................................... 25
2.3 Methods ......................................................................................................................................... 29
2.3.1 Test Configuration .................................................................................................................. 29
2.3.2 Markers .................................................................................................................................... 32
2.3.3 Data Collection and Processing ........................................................................................... 34
2.3.4 Static Analysis......................................................................................................................... 35
2.3.5 Dynamic Analysis ................................................................................................................... 35
2.3.6 Statistical Analysis.................................................................................................................. 36
ix

2.3.7 Human Subject Data .............................................................................................................. 36
2.4 Results ........................................................................................................................................... 37
2.4.1 Static Analysis......................................................................................................................... 37
2.4.2 Dynamic Analysis ................................................................................................................... 39
2.4.3 Human Subject Analysis ....................................................................................................... 39
2.5 Discussion..................................................................................................................................... 40
2.6 Conflict of Interest Statement .................................................................................................. 43
2.7 Acknowledgements .................................................................................................................... 43
3. UNDERSTANDING DISPLACEMENTS AND STRAINS OF THE GEL LINER FOR
BELOW KNEE PROSTHETIC USERS ................................................................................................ 44
3.1 Abstract.......................................................................................................................................... 45
3.2 Introduction................................................................................................................................... 45
3.3 Methods ......................................................................................................................................... 47
3.3.1 Participants .............................................................................................................................. 48
3.3.2 Prosthetic Componentry ........................................................................................................ 48
3.3.3 Test Procedure ....................................................................................................................... 48
3.3.3.1 Comparison of Devices .................................................................................................. 49
3.3.3.2 Measurement of Limb Motion within the Socket ........................................................ 50
3.3.4 Analysis .................................................................................................................................... 51
3.3.4.1 Relative Motion ................................................................................................................ 52
3.3.5 Statistical Method ................................................................................................................... 54
3.4 Results ........................................................................................................................................... 54
3.4.1 Participants .............................................................................................................................. 54
3.4.2 Displacements on the Residual Limb .................................................................................. 55
3.4.3 Strains on the Residual Limb................................................................................................ 55
3.4.4 Displacements Relative to the Prosthetic Socket .............................................................. 56
3.4.5 Angular Rotation of Limb Relative to the Socket ............................................................... 57
3.5 Discussion..................................................................................................................................... 58
3.6 Acknowledgements .................................................................................................................... 61
3.7 Funding .......................................................................................................................................... 61
4. INSTRUMENTED TRANSTIBIAL SOCKET FOR EVALUATING SHEAR AND NORMAL
FORCE: A CASE STUDY ...................................................................................................................... 62
4.1 Introduction................................................................................................................................... 63
4.2 Methods ......................................................................................................................................... 65
4.3 Results ........................................................................................................................................... 69
4.4 Discussion..................................................................................................................................... 72
5. FINITE ELEMENT ANALYSIS OF THE SOCKET TO LIMB INTERFACE WITH
EXPERIMENTAL DATA INPUTS ......................................................................................................... 75
5.1 Introduction................................................................................................................................... 76
5.1.1 Deep Tissue Finite Element Analysis (FEA) in Amputees ............................................... 77
5.1.2 Recent FEA Using FEBio ...................................................................................................... 77
5.1.3 Current Need for FEA in Prosthetics ................................................................................... 78
5.2 Methods ......................................................................................................................................... 79
5.2.1 Simplified One and Two Layer Models ............................................................................... 79
5.2.2 Realistic Four Layer Models ................................................................................................. 80
x

5.2.2.1 Mechanical Properties .................................................................................................... 81
5.2.2.2 Modeling Region of Interest and Geometry for Realistic Model .............................. 81
5.2.2.3 Loading and Boundary Conditions for Realistic Model ............................................. 83
5.2.2.4 Layer Contact Definitions for Realistic Model ............................................................. 83
5.2.2.5 Liner .................................................................................................................................. 84
5.3 Results ........................................................................................................................................... 86
5.3.1 Simplified Model Results ....................................................................................................... 86
5.3.1.1 One and Two Layer Models .......................................................................................... 86
5.3.2 Realistic Four Layer Model Results ..................................................................................... 90
5.3.2.1 Cylindrical versus Rectangular ..................................................................................... 90
5.3.2.2 Vertical Face Constraints ............................................................................................... 93
5.3.2.3 Slip versus No Slip Model Results ............................................................................... 94
5.3.2.4 Liner Stiffness Results.................................................................................................... 95
5.3.2.5 Liner Thickness Results ................................................................................................. 96
5.4 Discussion..................................................................................................................................... 97
5.4.1 Simplified Models ................................................................................................................... 98
5.4.2 Realistic Four Layer Models ................................................................................................. 98
5.4.2.1 Cylindrical versus Rectangular Geometry ................................................................... 98
5.4.2.2 Vertical Face Constraints ............................................................................................... 98
5.4.2.3 Slip versus No Slip .......................................................................................................... 99
5.4.2.4 Liner Stiffness ................................................................................................................ 100
5.4.2.5 Liner Thickness ............................................................................................................. 100
5.4.3 Engineering Model Relevance to Influence Clinical Practices ...................................... 101
5.4.4 Limitations, Future Work, Summary .................................................................................. 102
5.5 Conclusions ................................................................................................................................ 103
6.

CONCLUSIONS ............................................................................................................................. 104
6.1
Dissertation Conclusions .................................................................................................. 105

APPENDICES......................................................................................................................................... 109
APPENDIX A: FE Modeling Review ............................................................................................. 110
APPENDIX B: FE Modeling with Imaging Review .................................................................... 111
APPENDIX C: FE Modeling Using FEBio ................................................................................... 112
APPENDIX D: IRB Approval Letter .............................................................................................. 113
APPENDIX E: IRB Consent Form ................................................................................................. 114
BIBLIOGRAPHY .................................................................................................................................... 118

xi

LIST OF TABLES

Table 2-1. Displacement definitions. |12| indicates the distance between markers 1 and 2. The
left indicates the distances measured on the plaster limb and on the right the
deformable limb. ......................................................................................................34
Table 2-2. A. Comparison of caliper measured inter- marker distances to motion analysis data
for plaster limb replica with gel liner pin suspension interface. Measured distances
represent average +/- standard deviations of the three trials performed each day. The
difference between measured and caliper is the difference between measurement
method averages. No statistically significant differences were identified between
measured and caliper data.
B. Comparison of caliper measured inter-marker
distances to motion analysis data for deformable limb replica with gel liner pin
suspension interface. Measured distances represent average +/- standard deviations
of the three trials performed each day. The difference between measured and caliper
is the difference between measurement method averages. No statistically significant
differences were identified between measured and caliper data. .............................38
Table 2-3. Human subject data for displacements of marker locations beneath the prosthetic
socket. Standing data represents static weight bearing data. Walking ranges
represent the minimum to maximum displacement observed over five gait cycles. The
reported difference is the average and standard deviation of within trial dynamic
differences averaged across five gait cycles. ...........................................................40
Table 3-1. Participant characteristics. Residual limb length defined as inferior edge of patella to
distal end of the stump. Mobility grade scale classifies an individual’s ability to
ambulate or navigate their environment. (Gailey et al. 2002). Level K3 is defined as
the participant has the ability or potential for ambulation with variable cadence - a
typical community ambulator with the ability to traverse most environmental barriers
and may have vocational, therapeutic, or exercise activity that demands prosthetic use
beyond simple locomotion………………………………………………………………...55
Table 3-2. (a) Displacements for participants. Reported as average of five gait cycles plus or
minus standard deviation. (b) Strains for participants. Reported as average of five
gait cycles plus or minus standard deviation…………………………………………....56
Table 4-1. Force comparison across three conditions for the four particular periods during the gait
cycle…………………………………………………………………………………………71
Table 4-2. Initial contact and early stance statistical p values noting statistical differences reported
for normal force comparisons……………………………………………………………..71
Table 5-1. Mechanical properties of the model components selected for the FE models…..…..78
Table 5-2. Comparison results of two compressive pressures with FEA and theoretical axial stress
for a linear and non-linear response……………………………………………………...87

xii

Table 5-3. Compressive model results for a two layer neo-Hookean model with compressive
stress reported. Reported model outputs were evaluated in FEBio’s PostView
software………………………………………………………………………………….….88
Table 5-4. Compressive model results for a two layer neo-Hookean model with compressive
stress reported. Comparison averaged elements from the middle of layers which
resulted from model outputs for the top and bottom layers…………………………….89
Table 5-5. Combined loading (5 kPa distributed pressure and 6 mm displacement) results for
cylindrical versus rectangular four layer model………………………………………….91
Table 5-6. Results for cylindrical versus rectangular models with a 5 kPa distributed pressure on
top surface……………………………………………………………………………….….92
Table 5-7. Model with and without vertical constraints on the xz face in positive and negative y
directions. Middle locations were defined as elements selected in the center of the
layer’s cross sectional view at the layer surface as seen in Figure 5-9……………….93
Table 5-8: Comparison for slip versus no slip conditions in three locations (same liner used in
both). Location 1 and 2 are in the skin layer of the model and location 3 is at the muscle
to bone interface (Figure 5-14)…………………………………………………………….95
Table 5-9. Comparison of Von Mises, normal compressive and maximum shear stress for four
different modeled prosthetic gel liners with model comparisons across a no slip and
slip condition of the liner with respect to the skin………………………………….……96
Table 5-10. Comparison of TEC Urethane liner for three thicknesses of 9, 6 and 3 mm for no slip
and slip model definitions between the gel liner and skin interface………………..….97
Table A-1: Summary of FE modeling methodologies investigators implemented in the first decade
of prosthetic modeling. Original table, however, much of the content comes from a
published review article (Zachariah & Sanders 1996). E = Young’s modulus and v =
Poisson’s ratio……………………………………………………………………………. 110
Table B-1: Research summary FE models with improved imaging techniques but still using
linearly elastic material properties for biological tissues…………………………….. 111
Table C-1: Recent summary of FE models using FEBio in related biological research………..112

xiii

LIST OF FIGURES

Figure 1-1. Various prosthetic interface options available for patients with transtibial amputation
(Spires et al. 2014). .................................................................................................10
Figure 1-2. Transtibial suspension types (Spires et al. 2014) ....................................................11
Figure 1-3. Previous FE models for residual limb and prosthetic socket. Below knee (BK) and
above knee (AK) model examples (Zhang et al. 1998) ............................................21
Figure 2-1. Distal anterior tibia region skin break down and ulceration. .....................................26
Figure 2-2. a) Below knee prosthesis with pin suspension and gel liner. b) Clear Thermolyn
prosthesis developed for this research.....................................................................29
Figure 2-3. Experimental setup of replica limb and reflective markers beneath the socket with
pulley system application of force to the limb. Distal markers are closer to the pin
locking mechanism. .................................................................................................31
Figure 2-4. a) Marker placement over gel-liner interface for the plaster limb. b) Marker definitions
on the plaster limb. ..................................................................................................32
Figure 2-5. a) Replica of trans-tibial limb used in deformable limb testing. b) Marker placement
over gel-liner interface for the deformable limb. c) Marker definitions on the deformable
limb. .........................................................................................................................33
Figure 3-1. Example of clear thermoplastic duplicated socket (a, b) for one participant compared
with their original opaque prosthesis. Componentry and alignment was maintained
with only the socket exchanged for the experimental test configuration (a, b). The
duplicated socket allowed the cameras to track the markers within the socket whereas
the original socket does not……………………………………………………………….49
Figure 3-2. Calculation of displacements from anatomical marker locations on the gel liner within
the prosthetic socket……………………………………………………….………..……..50
Figure 3-3. Analysis of vertical displacement and rotation relative to the prosthetic socket during
a gait cycle. ∆ Z symbolizing the change in marker location for relative
displacement………………………………………………………………………………..53
Figure 3-4. Demonstration of distal gap between residual limb and prosthetic socket within the
socket just prior to initial contact and loading during walking…………………………..57
Figure 4-1. a) Top view of instrumented socket demonstrating flush internal curvature. b) Oblique
view of load cell securely affixed from the outside of the prosthetic socket……….....67
Figure 4-2. Assembled instrumented prosthesis as worn by the participant…………………..…67
Figure 4-3. Whole in gel liner to allow for contact and force measurement at the skin surface...68

xiv

Figure 4-4. Three conditions of normal and shear force during walking from heel contact to heel
contact on the right residual limb. Sashed lines in the same condition color represent
the ± standard deviation of force data as it was analyzed across 10 gait cycles….…70
Figure 5-1. a) Square neo-Hookean model b) Square neo-Hookean model with each layer divided
evenly in height c) Square neo-Hookean model with uneven layers……………..…..80
Figure 5-2. a) 3D geometry of four layer model b) Simplified geometry of FE model containing
bone, muscle, skin and gel liner constrained by experimental displacements and fixed
at the boundary of the bone. The gel liner to skin interface was either modeled as a
tied interface or a frictional sliding interface depending on the analysis to demonstrate
if slip occurred or not. A tied interface between muscle/skin used a connection of two
non-conforming meshes with a high penalty factor to prevent modeled tissue
separation. A rigid interface between bone/muscle was also defined to not allow
separation……………………………………………………………………………….…..82
Figure 5-3. Lower extremity model of bones and generalized soft tissue focused on an amputee’s
anatomy of limited limb length. The anatomical area enclosed by the rectangle
represents the fibular region of interest for which tissue thicknesses were modeled
based on average tissue thickness in this region……………………………………….82
Figure 5-4: a) No slip model and b) Slip model demonstrating contact differences between the
skin and gel liner interaction. As can be seen on the right, there is x direction translation
between the gel liner and skin……………………………………………………………..84
Figure 5-5. Tensile testing results for various prosthetic liners fell into four groups represented
by the liner boxed regions with decreasing stiffness (Daly & Odland 1979; Sanders et
al. 2004)……………………………………………………………………………………..85
Figure 5-6. a) 10 Pa compressive load b) 5 kPa compressive load; both modeled as a single
layer neo-Hookean material model but note linear and non-linear response
differences. (Scales are in Pa.)…………………………………………………………..87
Figure 5-7. Compressive 5 kPa distributed pressure alone with tied interface between layers.
Bottom layer E = 50kPa; Top layer E = 50kPa (Equal thickness layers; White dashed
line represents contact interface between materials; Scales are in Pa)……………...88
Figure 5-8. Compressive 5 kPa distributed pressure alone with tied interface between layers.
Bottom layer E = 50kPa (0.8m); Top layer E = 40kPa (0.2m). (White dashed line
represents contact interface between materials; Scales are in Pa)……………….….89
Figure 5-9. Four averaged elements were selected for the top surface of each layer. The same
placement of elements represented were consistent with deeper tissue layers….….90
Figure 5-10. Axis Symmetric Comparison (10 cm x 10 cm Cube versus a cylinder with a radius
of 5.6419 cm; A = 0.01 m^2). Combined loading with 5kPa pressure and 6mm x
direction displacement. Tied interface with no slip condition and Willow wood Alpha
Silicone Gel Liner…………………………………………………………………………..91

xv

Figure 5-11. Axis Symmetric Comparison (10 cm x 10 cm Cube versus a cylinder with a radius
of 5.6419 cm; A = 0.01 m^2). Loading with 5kPa pressure alone. Tied interface with
no slip condition and Willow wood Alpha Silicone Gel Liner. (Scales are in Pa)………92
Figure 5-12. a) Model with vertical constraints on the xz face in positive and negative y directions.
b) Model with vertical constraints removed. (Scales are in Pa.)………………………93
Figure 5-13: a) Gel layer hidden and presenting the model without slip. Demonstrating two
locations of Von Mises stress analysis on the skin. b) Location 3 is in the middle y
direction of the muscle to bone interface. (Scales are in Pa.) …………………...…..94
Figure 5-14. Gel layer hidden with difference is stress distribution in skin layer. a) Model with no
slip. b) Model with slip. (Scales are in Pa.)…………………………………………..…95

xvi

KEY TO ABBREVIATIONS

CAD

Computer aided design

CAM

Computer aided modeling

CT

Computed tomography

FE

Finite element

FEA

Finite element analysis

FEBio

Finite Elements for Biomechanics

FEM

Finite element methods

kPa

Kilopascals

LCS

Local coordinate system

MRI

Magnetic resonance imaging

OLCS

Origin of local coordinate system

UTS

Ultimate tensile strength

xvii

1. INTRODUCTION

1

1.1 Overview
Prosthetics is a clinical field in need of further investigation for the improvement of patient
care due to the complex nature of device development. Engineering principles can be used in
collaboration with clinical expertise to quantify key mechanical issues occurring at the residual
limb to prosthetic socket interface. Deep penetrating wounds in the skin called ulcers can form
on the residual limb within the socket. The formation of ulcers is not understood as the interface
mechanics have not been well researched.
Previous research has aimed to address certain portions of prosthetics which can
influence improvements of prosthetic design, specifically alignment and fit. However, numerous
gaps still exist in the basic knowledge of what boundary conditions are occurring at the socket to
limb interface. Quantitative data on limb motion within the socket, motion relative to the socket,
shear forces at the interface, and propagation of these loads to the skin level and deeper tissues
are all lacking in current literature.
The goal of this research was to study limb interface mechanics of the prosthetic/limb
boundary for better understanding of pressure ulcer formation. This work includes the following
aims:
Specific Aim 1: Develop and validate a quantitative method for assessing motions between the
prosthetic device and gel liner on the residual limb for patients with transtibial amputation.
Method: Transtibial rigid and deformable replica models were used in the development of a novel
method to validate measurement of motion capture thin-disc markers beneath the surface of a
clear prosthetic socket.

Markers were placed beneath the surface of the clear transparent

prosthetic socket and used to measure inter-marker distances which were compared statistically
to caliper inter-marker distance measurements.

2

Specific Aim 2: Determine limb displacements, strains, relative socket to limb displacements and
angular rotations of transtibial limbs within a prosthetic socket during gait.
Method: Eight participants consisting of nine limbs with below knee amputation were recruited
for the study. Reflective markers were placed on bony and soft tissue anatomical landmarks
throughout the residual limb beneath the surface of a clear duplicated prosthetic socket. A motion
capture system was used to track the markers. Displacements and strains were analyzed during
walking.

Anatomical locations were additionally used to compute relative limb to socket

displacements and relative rotation of the limb within the socket.
Specific Aim 3: Experimentally quantify longitudinal shear force and normal force within the
prosthetic socket for a transtibial amputee during walking in a single case study.
Method: A single transtibial prosthetic socket was instrumented with a two axis load cell to
measure forces at the internal socket wall. The participant walked in three conditions: gel liner,
three ply sock and a hole cut into the liner to measure forces at the skin. Force data located at
the mid fibular region of the residual limb was desired for constraining finite element models and
to further investigate interface mechanics in a pin locking suspension with gel liner interface.
Specific Aim 4: Determine the level of tissue stresses within a layered model including gel liner
interactions, constrained with experimental conditions of displacement and normal force.
Method: Numerically model the tissue layers in transtibial amputees with Finite Element Methods
(FEM) in Finite Elements for Biomechanics (FEBio). The residual limb section was defined as a
multi-layer model with different material properties for each layer based on literature. Layers
included bone, muscle, skin and the prosthetic gel liner. Non-linear hyperelastic materials were
used for the muscle and skin. This model simulated tissue layer stresses within the residual limb
in response to boundary conditions occurring at the socket interface. Boundary conditions used
were from experimentally collected displacement and force data.
3

This document has been formatted with a literature review plus four chapters, one
chapter per aim, followed by a conclusion chapter. Each chapter has been formatted as a
publication. A list of chapters can be seen as follows:
Chapter 1 Literature Review
Chapter 2 A New Method to Quantify Residual Limb Motion within a Prosthetic Socket
for Below Knee Amputees
Chapter 3 Understanding Displacements and Strains of the Gel Liner for Below Knee
Prosthetic Users
Chapter 4 Instrumented Transtibial Socket for Evaluating Shear and Normal Force: A
Case Study
Chapter 5 Finite Element Analysis of the Socket to Limb Interface with Experimental
Data Inputs
Chapter 6 Conclusions

4

1.2 Limb Loss Statistics
In the United States, approximately 1.7 million people are living with limb loss (ZieglerGraham et al. 2008). Additionally, an estimated 185,000 new amputations are performed each
year (Owings & Kozak 1998), with 120,000 of these amputations non-traumatic in nature
(Armstrong et al. 1997). Fifty-four percent of amputations are a result of complications from
vascular diseases such as diabetes, followed by 45% due to traumatic amputation (ZieglerGraham et al. 2008). Additionally with this many individuals undergoing amputation each year,
further amputation due to infection or ulceration particularly in the vascular patients is a clinical
concern. Progression to higher level amputations occurred 35% of the time when patients had
originally undergone a foot or ankle amputation (Dillingham et al. 2005). Furthermore, diabetic
amputees had a higher frequency of progression to a higher amputation than those nondiabetic
amputees (Dillingham et al. 2005). When analyzing racial differences in amputee rates, African
Americans exhibited greater risk for amputation, reportedly ranging from two to four times more
likely than Caucasians (Collins et al. 2002; Dillingham et al. 2005). Those patients who underwent
revascularization prior to the necessity for amputation were more often Caucasian, with elderly
African Americans receiving care at a significantly lower rate than Caucasians (Holman et al.
2012). In a study focused on Oklahoma Indians, the incidence rate of lower extremity amputation
was 1.8% of the population each year with males being twice as frequent as females with
significant co-morbidity being diabetes (Bahr et al. 1993). Regardless of race, gender or comorbidities, if limb health is not maintained, the 5-year rate of mortality for patients with lower
extremity amputation can be as high as 74% (Robbins et al. 2008).

1.3 Skin Disorders in Amputees
Persistent dermatologic concerns in amputees can restrict typical use of a prosthesis.
Maintaining a healthy residual limb is essential for preventing further amputation or pressure ulcer

5

formation.

Studies have investigated the range of skin disorders seen in a population of

amputees. In a sampling of patients with amputation, 34% experienced a skin problem including
epidermoid cysts, follicular hyperkeratoses, verrucous hyperplasia, calluses, ulcers, bacterial
folliculitis, tinea infection, eczema, dermatitis, transient erythema caused by friction, or
unexplained rashes (Lyon et al. 2000). These dermatologic problems can be classified by:
physical effects of wearing their prosthesis, allergic contact dermatitis, infection and constitutional
skin disease. Type of prosthetic design can also contribute to the frequency of skin disorders with
significantly more patients experiencing issues in soft socket prostheses (Koc et al. 2008) as
compared to those using silicon prostheses, regardless of suspension type. In the prostheticuser population investigated by Koc et al., 74% of patients experienced a skin problem and of the
142 patients enrolled in the study, the most common level of amputation was transtibial (n = 113).
Any of these skin complications can limit the patient’s use of their prosthesis, cause more serious
irritation or complications and potentially lead to reasons for further amputation.

1.4 Pressure Ulcer Statistics
Pressure ulcers are regional tissue damage areas from habitual excessive loading on the
skin from various body-interface conditions. Ulcers can be either superficial or deep in nature
depending on the loading conditions (Mak et al. 2010). Superficial pressure ulcers often result
from primarily frictional and abrasive rubbing of the skin relative to the prosthetic device. Deep
ulcers originate within a close proximity to bony prominences which can become massive lesions
from within before appearing at the surface. These bony prominences are particularly relevant
and troublesome in lower extremity amputees where loading is concentrated within the prosthetic
socket.

During daily activities, amputees wearing prosthetic devices experience high loads

between the prosthetic socket and the soft tissue around the residual limb; however, few studies
have actually quantified this loading.

6

In a systematic review, primary patient risk factors for pressure ulcer development were
identified as patient activity level/mobility, blood perfusion (including diagnoses such as diabetes),
and status of existing ulcers on skin (Coleman et al. 2013). Mobility factors include sub-categories
of bedfast, chair fast, walking with limitations and walking with no limitations (Coleman et al. 2013).
The most prevalent site on the lower limb is the posterior aspect of the heel, which unhealed heel
ulcerations commonly lead to amputation (Arao et al. 2013).

Important to note are the

morphological characteristics in skin when pressure ulcers develop. The tissue becomes less
tolerant to ischaemia and therefore less resilient to increased external forces (Arao et al. 2013).
The role of skin blood flow dynamics are key in understanding pressure ulcers because blood
flow function determines the ability of skin to respond to ischemic stress (Liao et al. 2013). Due
to reduced blood flow, the tissues of the extremities cannot receive adequate oxygen and
nutrients from the blood stream. Necrosis of the tissue begins, and infections often result (Bouten
et al. 2003). Reduction of blood flow can be caused by different changes in load applied to the
skin and has been proven to differ under normal or shear loads (Manorama et al. 2010).
Transcutaneous oxygen and blood perfusion levels decreased when shear loads were applied in
addition to normal loads (Manorama et al. 2010; Manorama et al. 2013). Understanding of load
on the skin, such as that due to a prosthetic, is important to understanding and preventing ulcer
formation. Particularly in the case of amputees, blood flow dynamics can be compromised due
to the loading conditions at the socket to limb interface. Ultimately, pressure ulcers are a great
clinical concern, increasing infection and leading to additional amputations.

1.5 Peripheral Arterial Disease: Risk Factors for Foot Ulceration and
Implications in Diabetic Patients
1.5.1 Importance of Restructuring Blood Flow After Amputation
Peripheral arterial disease is prevalent in 20-30% of diabetic patients (Marso & Hiatt
2006). Patients with peripheral arterial disease and diabetes are at a higher risk of lower extremity
7

amputation than those without diabetes (Jude et al. 2001). Many considerations need to be taken
into account when amputation is indicated. The following review will outline factors that relate to
amputation including peripheral arterial disease, foot ulcerations and the importance of
understanding blood flow patterns (angiosomes).

All relate to reasons individuals have

amputations and lead to prosthetic use.
Lower extremity peripheral arterial disease (PAD) is commonly associated with increases
in morbidity.

In worst cases, 1-2% of patients need major amputation (Khan et al. 2014).

Epidemiological studies have confirmed an association between diabetes and increased
prevalence of peripheral arterial disease (Marso & Hiatt 2006). The vessels often involved in
diabetic PAD patients are the tibial vessels and the distribution of pathology is located more
distally than in patients with PAD (Haltmayer et al. 2001). The abnormal metabolic state that
coincides with diabetes directly contributes to the development of atherosclerosis in PAD patients
with an increase in vascular inflammation. Amputation as a result of PAD is common for patients
experiencing treatments that have been unsuccessful to control infection (Marso & Hiatt 2006).

1.5.2 Diabetic Foot
Chronic foot ulcers are a result of foot lesions commonly in diabetic patients which persist
and contain complications often due to infection. Over 85% of amputations are preceded by an
active foot ulcer as these conditions are closely inter-related in diabetes (Boulton 2008). Patients
with a combination of infection and ischemia who presented with foot ulcers were 90 times more
likely to undergo midfoot or higher amputation compared to those with better wound management
(Prompers et al. 2007).

If occlusion occurred, lengthened duration of blood occlusion was a

characteristic associated with poor prognosis for amputation (Fagundes et al. 2005).
Experimental modeling of engineering mechanics related to arteries has provided understanding
of arterial wall density, poisson ratio, compliance, internal volume, pulse wave velocity and wall
thickness (Langewouters et al. 1984); however, computational modeling that addresses impaired
8

patient populations are missing from the literature and are essential for further understanding of
ulcer formation.

1.5.3 Reasons for Amputation
When limb salvaging methods are no longer viable, amputation may be the best option for
preserving ambulation capabilities. For individuals who are being considered for amputation, a
few considerations must be assessed. If the patient has a low chance of ambulation post
amputation, a through-knee-amputation is often indicated to prevent long term knee flexion
contractures which may develop in prolonged seated postures (Brown et al. 2012). For those
patients who have the potential to ambulate, appropriate surgical procedures are considered for
ample blood flow to the residual limb as well as ease for fitting into a prosthesis (Brown et al.
2012). Regardless of care taken to preserve healthy blood flow to the residual limb, re-amputation
can be necessary if wound ulcerations form or persist (Brechow et al. 2013).

1.5.4 Angiosomes
Blood flow patterns are complex and especially important to understand during
amputation. Vascular surgery is complicated when preserving adequate nourishment to the
residual tissue.

Surgeons are mindful of vascular territories, also known as angiosomes.

Angiosomes are regional subdivisions of branching arteries that supply regions of tissue. These
areas need to be considered during vascular surgery, especially limb salvaging surgeries or in
the worst case, amputation.

Compromised blood flow to a residual limb can lead to poor

nourishment of the residual limb and could predispose tissue necrosis. One study investigated
angiosomes by using fresh cadavers where regions of the lower extremities could be dissected
to separate muscles while preserving blood vessel connections (Taylor & Pan 1998). This
allowed for specific knowledge of which muscle was supplied with blood regionally as the arteries
bifurcated. Taylor and Pan et al., showed that the anterior leg compartment muscles were
9

exclusively supplied by the anterior tibial artery. It was clinically an important finding because
with a common vessel supplying one muscle followed by another in a narrow passage, this group
of muscles is particularly vulnerable to ischemia because the compartment is highly constricted
and has few vascular connections. When amputation is indicated, careful consideration of the
angiosome regions should be utilized (Attinger et al. 2006). As noted earlier, even with careful
consideration of blood vessels, and successful surgery, in many prosthetic users, ulcers continue
to form. Thus, further consideration of the prosthetic socket design to account for patient specific
blood occlusion regional concerns may decrease ulcerations in regions where vessels could be
easily occluded over bony prominences.

1.6 Prosthetic Device Components: Fabrication Selections
Prosthetics are designed to mimic anatomical function. In order to accomplish this task
with a variety of clinical limb presentations, there are numerous componentry options and
fabrication methods. The wide selection for prosthetic interface and suspension options (Figure
1-1 and 1-2) creates a complex prescription process for developing and selecting a definitive
device for a patient.

Figure 1-1. Various prosthetic interface options available for patients with transtibial amputation
(Spires et al. 2014).

10

Figure 1-2. Transtibial suspension types (Spires et al. 2014)

1.6.1 Sockets
The design of a prosthetic socket is always patient specific. Within a transtibial socket,
controlling the limb motion is essential for minimizing skin friction which could lead to pressure
ulcers. Additionally, minimizing excessive motion of the residual limb relative to the socket is
important for optimizing gait efficiency (Gard 2006; Fergason & Smith 1999). Three main types
of transtibial socket types are: patella tendon bearing (PTB), total surface bearing, and hydrostatic
(Spires et al. 2014). The PTB socket design became popular in the late 1950s (Radcliffe & Foort
1961). In a PTB socket, the device is meant to bear loads through pressure tolerant areas such
as the gastrocnemius, anterior tibialis, medial tibial flare, lateral shaft of fibula and patella tendon;
while it relieves pressure from sensitive areas such as the tibial crest, fibular head, hamstrings
and distal ends of the tibia and fibula (Spires et al. 2014). Advantages of a PTB socket are that
it creates a triangulation to control rotation of the socket relative to the residual limb but this is
difficult to implement in short residual limbs. Generally, the PTB style socket is thought of as
11

widely successful for transtibial amputees with reportedly 90% of below knee amputees
functioning well with a PTB socket (Galdik 1955; Pirouzi et al. 2014). Secondly, the total surface
bearing socket design loads the entire residual limb and is typically donned over a liner (Spires et
al. 2014). The gel liner can provide cushioning and absorption of rotational and shear forces while
aiming to provide equal distribution of loads over the residual limb. Lastly, a hydrostatic suction
socket generally provides less motion of the residual limb within the socket because it elongates
the soft tissue to increase stiffness between the bone and soft tissues for augmented stability
during gait (Kahle 1999). The most commonly implemented socket type is a PTB or modified
PTB and therefore was the focus of this dissertation research.

1.6.2 Socket Interface
The socket interface with the residual limb ranges from direct skin contact to gel liner
interfaces. In a direct socket interface, individuals need to have a limb with adequate soft tissue
covering bony prominences because no barrier exists between the skin and hard socket (Spires
et al. 2014). For patients requiring a cushioned interface, soft materials comprising of either a
soft plastic flexible inner liner or foam inserts can be used within the socket (Spires et al. 2014).
These provide manufacturing adaptation capabilities for anatomical changes and sensitive areas
to improve comfort but they require daily cleaning as they can absorb perspiration. Lastly, a gel
liner typically made of urethane, silicone or thermoplastic elastomer can decrease friction and
shear forces against the skin and can also provide cushioning (Spires et al. 2014). Gel liners are
most commonly prescribed for persons with below knee amputations (Boutwell et al. 2012) and
therefore the following research includes the use of a gel liner.

1.6.3 Suspension
Numerous factors are taken into account when considering a suspension type, these
include skin condition, volume, limb length, available range of motion (intact knee health), patent’s
12

activity level, comorbidities (peripheral neuropathy, vascular disease, cardiac disease) and an
individual’s cognitive level (ability to maintain device) (Kapp 1999; Pritham 1979). Seven typical
suspension options include (Figure 1-2): joint and corset, sleeve, supracondylar cuff,
supracondylar suprapatellar, supracondylar, gel liner with pin, and subatmospheric (Spires et al.
2014).
First, joint and corset suspension provides an above the knee brace system to help with
knee medial/lateral instability with an increased weight-bearing surface. However, this adds
weight to the prosthetic device with potential for increased pistoning of the residual limb relative
to the socket. Pistoning is when the prosthetic device translates vertically with respect to the
residual limb due to the suspension of the limb (H. Gholizadeh et al. 2014). Further disadvantage
is potential thigh musculature atrophy from not using the muscles for stability (Spires et al. 2014).
Secondly, sleeve suspension consists of a frictionous tightly fit outer sleeve that
encompasses the socket with the above knee residual limb. It is excellent for patients with a long
residual limb, stable knee ligament structures, good hygiene and no vascular comorbidities. Yet,
sleeve suspension can restrict knee motion, causing skin problems and it can be difficult to don.
The next set of suspension types (supracondylar cuff, supracondylar suprapatellar, and
supracondylar) function based on the ability to use bony anatomic landmarks for suspension and
occasionally additional straps. They are generally easy to don, provide increased medial/lateral
stability by crossing the knee joint but can be difficult to fit in obese patients (Spires et al. 2014).
Then there are gel liners with pin suspension set ups. These utilize a locking mechanism
on the distal end of the gel liner which attaches to the socket to hold the liner inside the socket
(Fergason & Smith 1999). The gel liner provides a frictional interface to suspend the complete
device off of the residual limb. A gel liner with pin suspension allows for less restriction of knee

13

range of motion, absorption of rotational forces, allows for volume fluctuations by adding socks
within the socket, and added cushioning within the socket (Spires et al. 2014).
Lastly, subatmospheric pressure suspension provides excellent suspension with
decreased pistoning at the residual limb, added proprioception and shown to help wound healing
(Brunelli et al. 2009). Major disadvantages to a subatmospheric pressure suspension though are
high maintenance for adequate function with expensive equipment cost and added weight to the
prosthesis.

The most frequently implemented suspension type consists of the pin locking

mechanism with liner interface and therefore was a focus when designing this dissertation
research in a prospective experimental study.

1.6.4 Combinations of Socket Interface and Suspension
A recent worldwide systematic review concluded there is no singular clinical standard for
suspension methods in transtibial amputees; however, the most favored setup by users consisted
of the total surface bearing socket with gel liner interface pin/lock suspension system (H.
Gholizadeh et al. 2014). The implementation of soft gel/silicone inner layers with pin locks has
greatly improved the function of artificial limbs by allowing a more comfortable prosthetic solution
with greater movement at the proximal joint (Heim et al. 1997). Notable however is the large
friction between the gel or silicone interface and residual limb skin which is clinically stated to
reduce the pistoning motion when the artificial foot contacts the ground (Narita, Yokogushi, Shii,
Kakizawa & Nosaka 1997). While this friction interface stability allows for functional clinical
benefits, understanding the skin surface during ambulation is important as it may be the source
of tissue breakdown which has become covered by the gel liner. Therefore, liner motion should
be monitored as it has become a common and preferred method of suspension interface in
transtibial amputees.

14

1.7 Experimental Literature on Lower Extremity Amputees
1.7.1 Interface Pressures
Numerous studies have experimentally assessed the pressure distribution within the
socket (Muller & Hettinger 1952). Key studies have documented pressure distributions for many
years, using this technology to assess stump-socket pressure and comparisons between different
socket types (Pearson 1974; Meier 1973; Leavitt et al. 1970; Appoldt et al. 1969; Naeff & van
Pijkeren 1980; van Pijkeren et al. 1980). For example, pressure distributions during self-selected
over ground flat walking are not predictable of pressures during walking on stairs, slopes and
uneven ground (Dou et al. 2006). Regardless of task, regional pressure differences can be
observed over the residual limb when comparing the regions of the patellar tendon, lateral tibia,
medial tibia, anterodistal tibia and popliteal depression. Highest pressures were observed in
normal gait at the popliteal depression (back of the knee joint), followed by the anterodistal tibia
and patellar tendon (Dou et al. 2006). Most recently, pressure distribution at the socket to limb
interface has been implemented to compare differences in prosthetic componentry, suspension
types, and liners (Beil et al. 2002; Hossein Gholizadeh et al. 2014; Ali, Abu Osman, et al. 2012;
Wolf et al. 2009; Boutwell et al. 2012; Eshraghi et al. 2013; Gholizadeh et al. 2015; Ali, Osman,
et al. 2012).
One study followed a single patient and found that the anterior distal residual limb peak
pressures were almost 10 times higher in the patellar tendon bearing socket and the patient
reported increased comfort in the total surface bearing socket (Gholizadeh et al. 2015). For
another study with twelve unilateral transtibial amputees, greatest peak pressures occurred at the
mid-posterior location (Wolf et al. 2009; Ali et al. 2013). These regions however were not
consistent with Dou et al., where in stair climbing they found notable changes when compared to
walking in the anterior and proximal areas above the patellar tendon region.

15

Prosthetic liner differences can also influence limb comfort and perceived pressure
distribution within the socket. Liners provide a layer of cushion between the limb and the socket;
depending on the limb architecture, more padding over bony landmarks may be desired to reduce
high peak pressures. In one study, fibular head peak pressures were significantly reduced with
a thicker liner, and resulted in increased patient comfort (Boutwell et al. 2012). Further mechanics
based research needs to be conducted to address the correlation of liner selection with increased
risk of pressure ulcer development.

1.7.2 Shear Stress at Interface
Pressure distribution is describing the compressive nature of the limb tissue, however, a
combination of normal and shear stresses are more valuable for describing tissue break down
due to skin blanching and blood occlusion. First published in 1992, the development of strain
based transducers established a method to measure shear stresses in two orthogonal directions
on the plane flush with the inside of the socket (Sanders et al. 1992). Three participants were
recruited for the study in which custom total-contact patellar-tendon-bearing prosthesis were
designed and fabricated for each transtibial amputee. Each socket was lined with a Pelite
interface which was custom designed to fit without the use of an additional sock or nylon sheath
and was suspended by a latex sleeve. This allowed for shear stresses to be measured directly
at the skin surface.
To further the understanding of shear stresses at the surface of the residual limb, Sanders
et al. improved on the original work expanding shear stress measurement to thirteen locations on
two patients with transtibial amputation (Sanders et al. 1997). Pressures as well as resultant
shear stress maxima were recorded during gait and the resulting timings of these loads. Areas
of highest shear stress were consistently at the anterior distal location (Sanders et al. 1997).
Timings of when maxima resultant shear stresses occurred were variable (Sanders et al. 1997).

16

This socket interface measurement was implemented in numerous studies to quantify differences
in shear stresses over time at daily and six month time points as well as changes due to prosthetic
alignment or various prosthetic componentry options (J. E. Sanders et al. 1998; Sanders et al.
2000; Sanders et al. 2005). Most recently Schiff et al. instrumented load cells into a transtibial
socket to further explore shear forces for amputees with and without distal tibia-fibular bone
bridges (Schiff et al. 2014). Further understanding of load transfer between the residual limb and
the prosthetic socket is not only important in ulcer formation but also in surgical decision making
of best amputation practices (Schiff et al. 2014). Extensive shear force information is currently
lacking, more experimental work combined with finite element modeling can further this
understanding of shear throughout the limb and implications loading on deep tissue stresses.

1.7.3 Friction and Prosthetic Liners
In order to better understand the interaction of prosthetic gel liners at the residual limb
interface, detailed descriptions of the material properties of commonly used interface materials
have been researched (Emrich & Slater 1998; J E Sanders et al. 1998; Sanders et al. 2004). It
was discovered that normal force versus shear force curves were nonlinear and the coefficient of
friction increased with higher applied force (J E Sanders et al. 1998). Later, the Sanders group
improved upon their previous work testing 15 products for classification of material performance
under compressive, frictional, shear and tensile loading conditions (Sanders et al. 2004).
Understanding liner materials is essential in optimizing prosthetic fits, tailoring prosthetic needs
based on a patient’s limb structure and preventing ulcer formation. These data in conjunction
with a finite element model optimizing gel liner materials will be helpful for understanding interface
forces and movements between different interfaces of the skin and liner or the liner and prosthetic
socket.

17

1.7.4 Pylon to Socket Interface: Load Cells, Forces, Moments, Inverse Dynamics
Load cells have been used in prosthetic research to instrument the junction of the
prosthetic pylon to the base of the prosthetic socket. Neumann et al. instrumented a load cell at
the base of the prosthetic socket in a portable manner to test three below knee amputees walking
on various terrains and curved pathways to measure real world situations of transverse plane
moments within the pylon (Neumann et al. 2013). Their purpose of analyzing this planar moment
was to hypothesize the contributing factors within the socket to these transverse plane moments.
The transverse moments represent when the residual limb and socket are attempting to rotate
relative to each other possibly generating shear forces at contact points within the socket. Higher
reported transverse moments were recorded when patients were asked to walk in a curved
circular path (Neumann et al. 2013). Similar studies instrumented load cells for above knee
amputees to measure forces and moments for multi-body simulation and inverse dynamics during
gait (Dumas et al. 2009; Schwarze et al. 2013). Schwarze et al., successfully validated a multibody simulation for calculating loads on the prosthesis interface for above-knee amputees.
However, none of these data directly assess loads and boundary conditions occurring at the limb
to socket interface for understanding localized regions of common tissue breakdown. These data
are necessary to fully model the liner to skin to device interface.

1.7.5 Kinematics: Whole body and Limb within Socket
Numerous experimental studies have investigated aspects of kinematic changes of
amputee gait in a whole body analysis as well as within socket kinematics. One approach used
stereogrammetric analysis to quantify skeleton relative to socket motion and skin strain during
strenuous motions such as a sudden stop and stepping down from stairs (Papaioannou et al.
2010). Roentgenological technology has been used to quantify movement between the stump
and socket (Erikson & Lemperg 1969). However, these analyses were limited to a small imaging
view due to the instrumentation. Key findings included maximum relative strain of proximally

18

located markers to be 8-10%, which is important for clinicians to know for optimizing prosthetic
design fit (Papaioannou et al. 2010). Before this more complex experimental study had been
developed for dynamic use, x-rays were initially used to quantify static positioning of the residual
limb bone structure within the socket (Friberg 1984; Newton et al. 1988; Lilja et al. 1993).
Ultrasound has also been used to measure planar motion of the femur relative to socket
in trans-femoral patients (Convery & Murray 2001; Convery & Murray 2000). The ultrasound
technique was compared to x-ray methods of determining frontal and sagittal plane angles of the
femur relative to socket and results were inconclusive (Convery & Murray 2001; Convery & Murray
2000). Attempts to use ultrasound were creative however, this application would not work due to
the presence of the tibia and fibula. Alternative experimental methods need to be developed to
quantify limb motion within the socket.
Dynamic analysis of socket relative to limb motion during walking is important to measure
for improving prosthetic device fit.

A noncontact sensor was developed from a lightweight

photoelectric sensor positioned beneath the socket to assess pistoning within the socket (Sanders
et al. 2006). Displacements during swing relative to stance phases of gait were obtained, proximal
displacements averaged 41.7 mm across multiple gait cycles for a single transtibial amputee
(Sanders et al. 2006). Motion was greater than expected of the socket relative to the residual
limb and therefore further exploration should quantify this in more patients and more regions
within the prosthetic limb.
Lastly, motion capture has been utilized to quantify limb motion within the socket (Childers
& Siebert 2015; Gholizadeh et al. 2012). Gholizadeh et al. assessed vertical displacement of the
limb relative to the socket along the lateral aspect of the residuum for two different liners. However
this was not conducted during walking but rather a progression of full-weight bearing, semi-weight
bearing, non-weight bearing and with 30, 60 or 90 N loads. Key findings compared the two liners

19

demonstrating the Seal-In X5 liner decreased pistoning by 71% compared to the Iceross Dermo
liner (Gholizadeh et al. 2012). Another study drilled holes in the prosthetic socket of a single
subject to allow for three motion capture markers to be placed on the gel liner which extended out
from the prosthetic socket (Childers & Siebert 2015). Residual limb movement relative to the
prosthetic socket demonstrated about 5 mm differences proximally versus distally; however this
only represented one person, and three locations with respect to the prosthetic socket (Childers
& Siebert 2015). The limb may be moving in uneven displacements depending on the soft tissue
or bony anatomical structure; therefore, further investigation should evaluate motion capture
methods in more regions of the residual limb during walking.

1.7.6 Residual Limb Volume Changes
Throughout an amputee’s life, management of their residual limb volume is essential for
maintaining proper socket fit and accounting for within day volume changes for the purpose of
minimizing pressure ulcers and wounds (Sanders & Fatone 2011).

Numerous reasons for

amputation exist, but regardless of initial etiology the residual limb during the first 12-18 months
changes considerably in shape, tissue structure and volume (Prosthetists 2004). Immature limbs,
just after amputation, undergo extensive edema and muscle atrophy, therefore socket volume
must be adjusted frequently (Golbranson et al. 1988). After this period of initial healing the limb
is then considered a mature limb; however, daily fluctuations in volume still occur and can often
be problematic (Zachariah et al. 2004; Sanders et al. 2009). Daily volume changes in mature
limbs are thought to be a product of pooling of blood in the venous compartment, arterial
vasodilatation and changes in interstitial fluid volume (Zachariah et al. 2004; Sanders et al. 2009).
The amount of daily volume fluctuation is also thought to be a function of comorbidities, prosthesis
fit, activity level, ambient conditions, body composition, dietary habits and for women, menstrual
cycle. A prosthetists’ role in volume management is essential for determining proper socket
design, prescription of within socket accommodations for volume fluctuation and determination of
20

the need for a new socket. Shape and volume changes in the residual limb are believed important
to changes in limb-socket interface pressure and shear stress distributions, which may in turn
lead to socket fit problems, including gait instability and skin breakdown (Sanders et al. 2005).

1.8 Finite Element Modeling of Residual Limb and Deep Tissue Injury
Finite element analysis and computer-aided design have improved upon a once purely
artisan field with increasing knowledge of within socket mechanical interactions, deep tissue
responses to loading, and improved socket design. First, in the late 1980s, FE modeling was
introduced as a potential instrument for prosthetic socket design (Krouskop et al. 1987). While
the model was simplistic, it was the first step towards future work aimed to address mechanics
within the socket that could not be seen with the eye. Other models quickly immerged using CT,
MRI and simplified geometries of amputee limbs with a generic layer of tissue, rigid bone structure
and an encompassing socket bound by external loads were implemented to investigate socket
interface mechanics (Brennan & Childress 1991; Reynolds & Lord 1992; Sanders & Daly 1993;
Steege & Childress 1988; Silver-Thorn 1991; Quesada & Skinner 1991; Zhang et al. 1995).

Figure 1-3. Previous FE models for residual limb and prosthetic socket. Below knee (BK) and
above knee (AK) model examples (Zhang et al. 1998)
21

While theoretically any model can be solved under the correct series of parameters, the value of
a FE model improves with experimental validation. During novel model development, comparison
with experimental data is essential to confirm a realistic clinical situation is being represented by
the model.
More recent models have refined tissue layers, interactions between those layers and
included mathematical constitutive models of elastic non-linear material properties to mimic
biological tissues. Real-time patient specific finite element analyses have been conducted to
assess internal stresses of soft tissue for the continued purpose of improving prosthetic fit
(Portnoy et al. 2007). The development of these models allowed for the investigation of deep
tissue injury when applied to loading of a transtibial’s limb (Portnoy et al. 2008; Portnoy, Siev-Ner,
Shabshin, et al. 2009; Portnoy, Siev-Ner, Yizhar, et al. 2009; Portnoy et al. 2010; Portnoy et al.
2011; Gefen et al. 2008). Key findings have discovered higher stresses accumulating at the
tissue-to-bone interface rather than more superficial tissue layers, indicating close monitoring of
limb health is essential in this clinical population. Pressure ulcers may be forming in the deep
layers of tissues far before they present at the skin surface.

Further detailed refinement of FE

models is needed to mathematically account for the limb’s complexity as well as customized
model inputs based on experimental data. Finally, validation of a FE approach with experimental
data will strengthen usage for predicting the onset of pressure ulcers and help to indicate the
prosthetic componentry that will minimize certain boundary conditions leading to limb wounds.

1.9 Conclusions
The medical complexity of blood flow dysfunction, prosthetic device design, experimental
methods in amputees and finite element models have been described in this literature review.
Many considerations need to be addressed when amputating an infected and poorly nourished
diabetic limb. Extensive research and collaboration with engineering should be considered to

22

understand the mechanical limitations of arterial wall strength, blood flow to peripheral regions
and remapping of blood flood pathways during amputation. Specific research methods to address
these topics in amputated limbs is necessary to define the primary factors which may lead to
ulceration on the residual limb from poor tissue nourishment.
With extensive componentry options for making transtibial prosthetic devices, there is not
one clinically perfect answer with a proven algorithm to have optimized performance using a
particular device. The clinical judgement and crafted skill of developing a prosthetic device is
extremely complex. However, experimental research will continue to contribute quantitative data
to the field of prosthetics to guide appropriate device development.
A variety of research studies have investigated crucial questions relevant to the transtibial
patient population focusing on prosthetic componentry, pressure distribution, walking kinematics,
within socket fit and forces measurement. However, in such a complex field with reoccurring
problems of pressure ulcers, more research is needed to address numerous unknowns. As
prosthetics transitions from an artisan field to an integrated computerized technical trade,
biomedical engineering research can be at the forefront of new methods, prosthetic designs and
improved quality of life for patients living with amputations.
As a result of the above literature review, experimental and modeling aims have been
developed to address areas of research needed to quantify limb motion beneath the surface of a
transtibial socket including measurement of displacements, strains and shear forces throughout
the residual limb with understanding of deep tissue response. These novel studies to address
these research questions have been outlined in the following chapters.

23

2. A NEW METHOD TO QUANTIFY RESIDUAL LIMB MOTION WITHIN A
PROSTHETIC SOCKET FOR BELOW KNEE AMPUTEES

24

2.1 Abstract
Many amputees who wear a leg prosthesis develop significant skin wounds, called
pressure ulcers, on their residual limb. The exact cause of these wounds is unclear as little work
had studied the interface between the prosthetic device and the limb. Our research objective was
to develop a quantitative method for assessing limb displacement patterns during walking for
patients with transtibial amputation. Using a reflective marker system and a custom clear socket,
three evaluations were conducted: 1) a clear transparent test socket mounted over a replica of a
patient’s leg cast in plaster, 2) a deformable leg model and 3) a patient’s leg. Using a motion
capture system, distances between markers were measured with a digital caliper and compared
with the motion capture system. Dynamic trials were then collected while the non-human limbs
were vertically displaced to measure changes in inter-marker distance due to vertical elongation
of the gel liner. Static inter-marker distances within day and across days confirmed the ability to
accurately capture displacements using this new approach. Furthermore, a single human subject
was tested with this approach and larger displacements were found distally as compared to
proximally. Uneven deformation is an interesting finding because clinicians may be
underestimating the displacement particularly at the distal end, where ulcers commonly occur.
These results encourage this novel method to be applied to a larger sample of amputee patients
during walking to assess displacements and the distribution of limb deformation within the socket.

2.2 Introduction
In the United States, approximately 1.7 million people are living with limb loss (ZieglerGraham et al. 2008). In addition, there are roughly 185,000 new amputations performed each
year (Owings & Kozak 1998). Fifty-four percent of amputations are a result of complications from
vascular diseases such as diabetes, followed by forty-five percent due to traumatic amputation
(Ziegler-Graham et al. 2008). Following amputation, wound management and healing is essential

25

because further amputation is a clinical concern due to infection or ulceration (Dillingham et al.
2005).
Pressure sores or ulcers are deep penetrating wounds that frequently occur on the
residual limb at the socket interface (Figure 2-1). These wounds are painful, prone to infection,
and disruptive to the patient (Mak et al. 2010). Numerous factors contribute to the formation of
pressure ulcers both in clinical comorbidities and anatomical residual limb architecture. A recent
systematic review concluded there is no singular risk factor to explain the occurrence of pressure
ulcers (Coleman et al. 2013).

Clinical comorbidities leading to ulcers often include poor

circulation, venous insufficiency, diabetes, kidney insufficiency, hypertension, lymphedema, and
inflammatory diseases (Nixon et al. 2006; Coleman et al. 2013).

Figure 2-1. Distal anterior tibia region skin break down and ulceration.

Wound healing throughout the clinical care process is essential and often challenging as
there are many complications. Clinical factors that contribute to ulcer formation include: time
since amputation, degree of tissue remodeling, presence of bony prominences close to the
surface of the skin, quality of socket fit, and distribution of forces on the limb (Koc et al. 2008;
26

Lyon et al. 2000). Wound infection and poor healing are the frequent culprits leading to reamputation. In the case of trans-tibial amputations, sixty percent of patients undergo transfemoral
amputation, contralateral limb amputation, or pass away (Dillingham et al. 2005; Robbins et al.
2008; Ziegler-Graham et al. 2008).
In order to maintain ambulation, patient-specific prosthetic devices are designed and
manufactured (Johnson & Davis 2014; Edwards 2000; Narita et al. 1997). Prosthetists can
connect the socket to the residual limb through numerous means of suspension. A recent review
concluded there is no singular standard for suspension method in transtibial amputees; however,
the most favored setup by users was the total surface bearing socket with gel liner interface
pin/lock suspension system (H. Gholizadeh et al. 2014). Prosthetists implement this type of
system because it provides a less restricted knee range of motion, is believed to decrease
pistoning of the residual limb, increases proprioception between the limb and socket, and can
provide cushioning (Kapp & Fergason 2004; Johnson & Davis 2014; Kapp 1999).
The implementation of soft gel/silicone inner layers with pin/locks has greatly improved
the function of artificial limbs by allowing a more comfortable prosthetic solution with greater
movement at the proximal joint (Heim et al. 1997). Notable however, is the large friction between
the gel or silicone interface and residual limb skin which is clinically stated to reduce the pistoning
motion (Narita et al. 1997). Pistoning is when the prosthetic device translates vertically with
respect to the residual limb due to the suspension of the limb (Gholizadeh et al. 2012). While this
high friction interface allows for clinical benefits, understanding its effects at the skin surface
during ambulation is important as this interface may be a source of tissue breakdown. Therefore,
a need exists to understand how the liner moves relative to the prosthetic device and relative to
the skin.
Skin movement and residual limb loading are ongoing throughout the day for prosthetic
users, this loading, coupled with deformation and strain on the skin plays a role on ulcer formation
27

(Gefen et al. 2008). In particular, for lower-leg prosthetic users, we hypothesize that there is
uneven displacement of the limb within the socket with larger displacements occurring distally.
Resulting from uneven displacements, non-uniform strains or slippage of one surface to another
may result in regions of high shear forces on the skin and deeper tissues. Other researchers
have shown that shear loads on the skin significantly reduce blood flow to the loaded region
(Manorama et al. 2010; Manorama et al. 2013). A reduction of blood flow, over a duration of time
leads to skin necrosis and formation of an ulcer (Bouten et al. 2003). Additionally, loads at the
skin level can lead to deeper tissue stresses, again causing tissue damage and ulcer formation
(Gefen et al. 2008). To study the relative movement between the skin and the prosthetic device
we must first develop a method that permits this assessment.
Previous research has provided limited information on the movement that occurs between
the socket and liner interface (Childers & Siebert 2015; Gholizadeh et al. 2012). Dynamic
roentgen stereogrammetric analysis was also utilized to assess skin movement within the socket
however this study was limited to small movements so the participate remained within the small
capture region of the 3D digitizer (Papaioannou et al. 2010). None of these methods have studied
the complete displacement field of limb motion within the socket. In particular, research is needed
to compare longitudinal and transverse displacements between the socket and the residual limb
during regular activities such as walking. In order to perform this comparison, first a method must
be developed.
Motions occurring within the socket are not measured by prosthetists because the
interface region is not visually accessible and a method for quantifying these motions is not readily
available. Thus, the objective of this work was to develop and validate a quantitative method for
assessing limb displacement patterns for patients with transtibial amputation while wearing a
prosthetic device. This method used a clear prosthetic socket and motion capture markers to
obtain displacement data sets.
28

2.3 Methods
For the development of this method, research was focused on transtibial prostheses with
a gel liner interface and pin locking suspension (Figure 2-2).

Figure 2-2. a) Below knee prosthesis with pin suspension and gel liner. b) Clear Thermolyn
prosthesis developed for this research.

2.3.1 Test Configuration
Two replicas of an amputee’s transtibial limb were used in the development of the
experimental method: 1) an anatomical replica made of rigid plaster and 2) a deformable limb
model which mimics the anatomical structure and deformable nature of a transtibial limb (Burner
et al. 2013; Dombroski et al. 2014). The deformable limb has material components representing
bones, soft tissue and skin (Burner et al. 2013; Dombroski et al. 2014).

29

Using standard clinical prosthetic socket manufacturing methods, appropriate sized
sockets were fabricated, custom to each unique limb model. The model limbs were casted,
positive plaster models were modified, sockets were pulled using thermoplastic, and anatomical
trim lines were cut and smoothed. To allow for measurement and capture of the limb motion and
anatomical landmarks, the socket was manufactured out of a clear material called Thermolyn.
Thermolyn is a transparent, thermoplastic material commonly used in fabrication of prosthetic
“test sockets” for clinical use leading up to a definitive prosthetic socket. The limb and socket with
pin locking mechanism were then rigidly affixed to a metal base through the linkage of a prosthetic
pylon. This provided a stable base of support for the test configuration (Figure 2-3).

30

Figure 2-3. Experimental setup of replica limb and reflective markers beneath the socket with
pulley system application of force to the limb. Distal markers are closer to the pin locking
mechanism.

31

2.3.2 Markers
The production of the clear sockets allowed reflective thin-disc motion capture markers to
be positioned inside on the gel liner directly beneath the surface of the transparent socket and
spherical markers on the outside of the socket so that static anatomical locations could be
measured. Anatomical locations of interest included: anterior tibial tuberosity, anterior tibial crest,
distal end of the tibia, lateral proximal fibular head, distal fibula, intersegmental locations along
the lateral fibula, medial tibial condyl, soft tissue medial limb border to distal end, and
gastrocnemius muscle soft tissue locations on posterior calf are shown on Figures 2-4 and 2-5
while the displacement measures are presented in Table 2-1.

Figure 2-4. a) Marker placement over gel-liner interface for the plaster limb. b) Marker
definitions on the plaster limb.
32

Figure 2-5. a) Replica of trans-tibial limb used in deformable limb testing. b) Marker placement
over gel-liner interface for the deformable limb. c) Marker definitions on the deformable limb.
33

Table 2-1. Displacement definitions. |12| indicates the distance between markers 1 and 2. The
left indicates the distances measured on the plaster limb and on the right the deformable limb.

The plaster limb model was tested first with an initial marker set-up consisting of sixteen
reflective thin-disc markers beneath the clear, transparent socket (Figure 2-4). Following data
collection and analysis from the plaster limb, the deformable limb model had marker placements
adjusted to improve the representation of transverse and longitudinal inter-marker distances.
Therefore, in the deformable limb seventeen reflective thin-disc markers were used (Figure 2-5).
Markers were positioned to represent bony anatomical landmarks because these locations are
related to clinically known pressure intolerant areas and soft tissue areas likely to show larger
deformation. Three external spherical reflective markers were placed on the rod attached within
the limb replica and an additional three markers were placed on the pylon to track motion of the
external prosthetic device relative to internal marker locations.

2.3.3 Data Collection and Processing
A twelve camera motion capture system (Vicon Motion Systems Ltd.; Oxford, UK)
recorded the locations of all markers in three-dimensional space; the system was calibrated
before each data collection. Accuracy of this motion capture system is +/- 0.5 mm for inter-marker
distance calculations.

Inter-marker distance calculations were conducted between pairs of
34

markers and a relative displacement field was calculated based on the locations of individual
markers moving relative to each other. Inter-marker distances were computed as the magnitude
of the vector between the two markers. Thus, |12| indicates the magnitude between markers one
and two. Reported inter-marker distance definitions on the limb relative to their anatomical
representation can be found in Table 2-1.

2.3.4 Static Analysis
For each limb model, rigid and deformable, three static motion capture trials were
collected.

In both limbs, inter-marker distances reported longitudinal and transverse limb

distances. Specific to each limb, these same anatomically defined inter-marker distances were
also measured with a digital caliper in millimeters to the hundredths place with measures being
repeated three times, and averaged. Thus comparisons could be made to see if the plastic device
caused distortion of marker distances with the motion capture system. For both limbs, this test
protocol was conducted on three separate days to test within day accuracy and across testing
day repeatability.

2.3.5 Dynamic Analysis
To ensure movement through the motion capture volume would not negatively influence
intersegmental dimensions, the plaster model was moved in biplanar motions. Multiple dynamic
trials simulated a walking motion through the capture volume and inter-marker distances were
computed and compared.
To simulate pistoning, a dynamic condition was created that displaced the limb upward.
The limb and socket with pin locking mechanism were rigidly affixed to a metal base through the
linkage of a prosthetic pylon. This provided a stable base of support for the test configuration
(Figure 2-3). A rope-pulley system was used to induce pistoning while the marker data were used
to compute inter-marker distances. The prosthetic socket was securely mounted via a pylon to a
35

steel base. This fixed approach was used to mimic the contact phase of the foot during walking.
Our hypothesis was that larger displacements would be observed distally on the residual limb due
to pistoning.

2.3.6 Statistical Analysis
For data collected within the same day, repeated measures t-tests (SPSS Statistical
Analytics, Armonk, New York) were performed to compare static motion capture marker distances
with caliper measured distances. Separate analyses were run for the rigid and deformable limbs.
One-way analysis of variance (ANOVA) statistical tests were performed for motion capture
distances measured across the three testing days.

Significant differences for all statistical

analyses were defined with a P value of 0.05 or less. Dynamic displacement distances were
reported regionally in both limb setups as an average maximum displacement over three cyclical
vertical displacements. Biplanar dynamic inter-marker distances were compared with static
motion capture inter-marker distances using a t-test (SPSS Stastical Analytics, Armonk, New
York) to statistically quantify measured differences through the clear socket when movement was
introduced.

2.3.7 Human Subject Data
Finally, our experimental method was implemented with a single human subject during a
walking cycle. This work was conducted under an approved Human Subject’s IRB protocol (#14089M).
A prosthetic device identical to the patient’s original was formed with the clear, transparent
Thermolyn. The patient’s limb length and circumferential dimensions were smaller than either the
plaster or deformable models. In order to avoid merging of markers, the placement was modified
to fourteen thin-disk reflective markers (Figure 2-5).

However, this setup still allowed

displacement measurements along the same anatomical regions.
36

Marker distances were

evaluated relative to the patient’s gait cycles (Miller 2009). The maximum and minimum intermarker distances were reported within a single gait cycle and five independent trials of one gait
cycle were each evaluated.

2.4 Results
2.4.1 Static Analysis
First reported are inter-marker distances for the plaster replica limb (Table 2-2A). Next,
data are reported for the static inter-marker distances for the deformable limb replica (Table 22B). The differences for the plaster limb between caliper and motion data ranged from the lowest
of 0.00 to the highest of 0.46 mm. Differences between caliper and motion capture for the
deformable limb ranged from 0.01 to 0.50 mm. No statistically significant differences were found
between the motion capture and caliper distances for the plaster leg with regard to the
measurements obtained using the different techniques. This was also true for the deformable
limb.

37

Table 2-2. A. Comparison of caliper measured inter- marker distances to motion analysis data
for plaster limb replica with gel liner pin suspension interface. Measured distances represent
average +/- standard deviations of the three trials performed each day. The difference between
measured and caliper is the difference between measurement method averages. No
statistically significant differences were identified between measured and caliper data. B.
Comparison of caliper measured inter-marker distances to motion analysis data for deformable
limb replica with gel liner pin suspension interface. Measured distances represent average +/standard deviations of the three trials performed each day. The difference between measured
and caliper is the difference between measurement method averages. No statistically
significant differences were identified between measured and caliper data.

Repeated measures paired t-tests revealed no significant differences for within day
measurements for either the plaster or deformable limb. Additionally, an ANOVA revealed no
statistically significant differences between caliper and motion capture data for either the plaster
or deformable limb across testing days.
38

2.4.2 Dynamic Analysis
Inter-maker distance and range of dynamic motion of the gel liner were also analyzed for
the rigid plaster and deformable amputee limb replicas.

These inter-marker distances are

reported in Table 2-2, and are the minimum and maximum range of displacement over five vertical
displacement trials. Dynamic ranges were greatest in distal and vertical inter-marker locations
for both limbs yet even greater displacements were observed in the deformable limb than the
plaster. The average applied load for vertical displacement was 119.7 Âą 23.1 N with resulting
range of 10-20 mm of displacement between the markers on the rod attached to the limb replica
and the pylon (Figure 2-3).
For biplanar dynamic motion of the model through the capture volume, statistical analysis
using a t-test compared static and dynamic inter-marker distances. No statistically significant
differences were found between the motion capture static trials and motion capture dynamic trials
for calculated distances using the plaster leg.

2.4.3 Human Subject Analysis
Dynamic measurement of inter-marker distances during walking for one patient resulted
in larger displacements distally versus proximally, with the largest regions of displacement at the
distal tibia and distal fibula (Table 2-3). Specifically, marker distances |45| and |89| which
represent the locations on the distal tibia and distal fibula showed displacements of 7.34-7.41 mm
longitudinally. Transverse marker distances corresponding to |12| and |23| which were along the
tibial tuberosity to the medial and lateral directions averaged displacements of 3.82-3.9 mm. The
human subject data reported displacements greater than observed in the non-human models with
human data exhibiting approximately 3-5 mm more displacement throughout the limb.

39

Table 2-3. Human subject data for displacements of marker locations beneath the prosthetic
socket. Standing data represents static weight bearing data. Walking ranges represent the
minimum to maximum displacement observed over five gait cycles. The reported difference is
the average and standard deviation of within trial dynamic differences averaged across five gait
cycles.

2.5 Discussion
In an effort to understand residual limb movement within the prosthetic, a method was
developed using thin disc markers located on the gel liner beneath clear plastic prosthetic sockets.
The measurement technique was validated by comparing motion capture data and digital caliper
data for static trials. Results showed that no statistically significant differences between measured
caliper inter-marker distances and those acquired by the motion capture system existed for either
the rigid or deformable limb models. Absence of statistically significant differences indicated that
motion capture system markers mounted beneath a clear thermoplastic prosthetic test socket
could accurately capture movement of the gel liner. Additionally, data were compared across
multiple sessions within a day, and across three days. No differences were found.
Limited data are available for quantification of limb motion within the socket (Childers &
Siebert 2015; Gholizadeh et al. 2012). Gholizadeh et al. assessed vertical displacement of the
limb relative to the socket along the lateral aspect of the residuum for two different liners. However
this was not conducted during walking but rather a progression of full-weight bearing, semi-weight
bearing, and non-weight bearing. Comparisons showed that liner type affected the magnitude of
pistoning; specifically, the Seal-In X5 liner decreased pistoning as compared to the Iceross Dermo

40

liner (Gholizadeh et al. 2012). Another study drilled holes in the prosthetic socket to allow for
three motion capture markers to be placed on the gel liner (Childers & Siebert 2015; Gholizadeh
et al. 2012). This work demonstrated approximately 5 mm differences proximally versus distally;
however the prosthetic device was altered, potentially affecting the fit and they were only able to
measure a few locations (Childers & Siebert 2015). Our new method allows for a much larger
evaluation of limb displacement measurements across the entire lower leg and this approach can
be used during dynamic activities such as walking, stair climbing or even running.
Our study also used a lab set-up with a plaster and deformable limb to emulate pistoning,
where the liner elongated beneath the rigid socket. Both the plaster and deformable limb models
showed displacement changes with larger displacements distally. Additionally, the deformable
limb yielded larger displacements than the rigid plaster limb. The plaster limb data set provided
insight into the liner response as there were no soft tissues below the liner. Based on the plaster
and deformable limb model results, it was hypothesized a human limb with soft tissue would
demonstrate even greater displacements than either the plaster or deformable limbs.
A clear prosthetic was manufactured for a single subject and our approach was used to
provide an initial indicator of what displacements would occur in a human data set. This initial
participant showed uneven displacement occurring within the prosthetic socket during walking.
The two highest regions of displacement were measured at the distal tibia and fibula. This result
was likely due to the combination of the soft tissue at the distal end of the leg and the pin/lock
suspension mechanism which “pulls” at the distal end due to the mass of the foot during swing.
The entire prosthesis is suspended from the distal pin on the gel liner, therefore it must carry the
load of the pylon/foot during swing. Increased movement at the distal end of the limb may lead
to higher friction at the skin surface leading to tissue irritation and the formation of ulcers.
Prosthetists are unable to measure these within socket displacements, therefore, this new method

41

will increase data available for prosthetists and will help support improved prosthetic device
design.
Based on these results, it is likely that the larger distal displacements will lead to higher
shear forces and stresses in deeper tissues. Once the motions of the gel liner relative to the
prosthetic device are quantified, then engineering principles can be used in a model to understand
the stresses, strains, deformations and stretch occurring on the skin and in deeper tissues. This
mechanics-based information, combined with clinical information, will allow us to understand the
local loading and provide further insight regarding localized tissue breakdown and ulcer formation.
Limitations of this study are a function of the type of prosthetic suspension being studied.
While gel liner interface with pin/lock mechanism suspension is a commonly used setup in
transtibial amputees, it does not allow for measurement directly at the surface of the skin. Without
the gel liner type of suspension method, there would be no stable form of suspension in the
prosthesis without extending the device above the knee (i.e. joint corset design). The large friction
between the gel liner and residual limb skin has been clinically stated to reduce the pistoning
motion (Narita et al. 1997). While this friction interface stability allows for functional clinical
benefits, understanding the skin surface during ambulation is important as it may be the source
of tissue breakdown which has become covered by the gel liner.
Clinically impact of this new method is high. Obtaining these quantitative data and
understanding the distribution and magnitude of displacements can potentially explain regions
where in incidence of pressure ulcer formation is high, commonly observed at the distal tibia and
distal fibula. Further, these data are likely to lead to improved patient-specific socket designs for
prosthetic devices that reduce local loads and minimize ulcerations on the residual limb. Future
research will apply this experimental method to more patients presenting with transtibial
amputations during walking to understand the displacement field occurring within the socket.

42

2.6 Conflict of Interest Statement
The authors do not have any conflicts of interest to disclose.

2.7 Acknowledgements
The authors would like to thank Patrick Logan, CPO and Leo DelCarmen for their support and
encouragement of the research. Also a special acknowledgement goes to the Orthotics and
Prosthetics School at Eastern Michigan University for the gracious use of their transtibial
deformable limb model without any financial involvement.

43

3. UNDERSTANDING DISPLACEMENTS AND STRAINS OF THE GEL LINER
FOR BELOW KNEE PROSTHETIC USERS

44

3.1 Abstract
Many people with amputation utilize a prosthetic device to maintain function and
ambulation. During the use of the prosthetic device, their residual limbs can develop wounds
called pressure ulcers. The formation of these wounds has been linked to deformation and loading
conditions of the skin and deeper tissues. Our research objective was to develop a complete
profile of displacements and strains of the residual limb within the socket during walking in
transtibial amputees. Seven regions within the limb were evaluated for calculated displacements
and strains in addition to six calculations of displacement and three rotations relative to the
prosthetic socket. Greatest displacements were observed in the distal region of the residual limb,
near the pin locking mechanism on the gel liner.

Calculated displacements were uneven

throughout the residual limb and reportedly greater than previous research, indicating a potential
problem with current prosthetic device design methods. This mechanics-based information,
combined with clinical information, will allow us to understand the local skin and muscle
displacements, the associated forces and will provide insights regarding localized tissue
breakdown.

Knowledge of limb movement within the prosthetic socket can help also help

prosthetists modify prosthesis design to reduce these displacements and strains.

3.2 Introduction
With improved safety mechanisms more people are surviving auto or military related
accidents and therefore living with a missing limb (Ziegler-Graham et al. 2008). In the United
States alone, 1.7 million individuals have had a foot or leg amputation; in general, 45% are due
to traumatic amputation and 54% are a result of complications from vascular diseases such as
diabetes (Ziegler-Graham et al. 2008).

In 35% of individuals with amputation, a second

amputation further up the leg occurs due to clinically unmanageable infections (Dillingham et al.
2005). Furthermore, diabetic amputees have higher frequency of secondary amputation than
nondiabetic amputees (Dillingham et al. 2005).
45

One cause for infection and a second amputation, are pressure ulcers. Pressure ulcers
are regional tissue damage caused by loading on the skin from body-device interface conditions,
particularly common over bony prominences (Le et al. 1984). Ulcers can be either superficial or
deep depending on the loading conditions (Mak et al. 2010). Common regions of ulcer formation
have been clinically noted to occur at the distal end of the tibial crest, lateral tibial flare, fibular
head and distal fibula (Dudek et al. 2006; Henrot et al. 2000).
Following transtibial amputation, a prosthetic socket is utilized to maintain ambulation for
patients. Understanding the limb to socket interface is essential for creating a well fit prosthetic
device and to minimize tissue damage and the potential for the development of pressure ulcers.
Pressure distribution analysis has been performed for over six decades, analyzing regions of high
and low pressure primarily during gait (Muller & Hettinger 1952; Gholizadeh et al. 2015).
However, pressure measurement usually requires thin-film sensors that suffer from errors with
changes in temperature, curvature, and humidity. Pressure also does not provide a complete
evaluation of loading conditions within the socket. To obtain a comprehensive understanding of
the limb-device interface, a better understanding of how the limb deforms and movement of the
limb with respect to the prosthesis is needed.
Understanding movement of the socket relative to limb during walking is important for
improving prosthetic device fit, however, few studies have tackled this problem. One study
created a noncontact sensor positioned beneath the socket to assess pistoning within the socket;
but this was limited to a single measurement at the distal end of the limb (Sanders et al. 2006).
Pistoning is when the prosthetic device translates vertically with respect to the residual limb due
to the suspension of the limb (H. Gholizadeh et al. 2014). Two other studies used motion capture
to quantify limb motion within the socket (Childers & Siebert 2015; Gholizadeh et al. 2012).
Gholizadeh et al. assessed vertical displacement of the limb relative to the socket along the lateral
aspect of the residuum for two different liners, however this was not conducted during walking but
46

rather for partial to full weight bearing positions. Another study drilled holes in the prosthetic
socket of a single subject to allow for three motion capture markers to be placed on the gel liner;
yet this did not represent a wide range of locations within the prosthetic socket (Childers & Siebert
2015). Uneven displacements are likely because of limb structure, soft tissue regions and the
walking movement. Therefore, further investigations are necessary to evaluate displacements
around the entire residual limb during dynamic daily activities such as walking.
Therefore, the objective of this work was to determine 1) limb displacements, 2) strains,
3) relative displacements between the limb and socket, and 4) angular rotations of transtibial limbs
within a prosthetic socket during walking. We hypothesized that larger localized displacements
and strains would occur in the distal region due to pistoning of the gel liner. We also believe this
pistoning will lead to higher shear forces on the skin and in deeper tissues influencing the stresses
making tissue damage more likely. Once the motions of the gel liner relative to the prosthetic
device are able to be quantified, then engineering principles can be used to model the resulting
stresses, strains, deformations, and stretch occurring on the skin and in deeper tissues. This
mechanics-based information, combined with clinical information, will allow us to understand the
local skin and muscle displacements, the associated forces and will provide insights regarding
localized tissue breakdown.

Further, these data are likely to lead to improved patient-specific

socket designs for prosthetic devices that reduce local loads and minimize ulcerations on the
residual limb.

3.3 Methods
The previously developed method to capture a displacement field of thin-disc reflective
markers beneath the surface of a prosthetic socket (Lenz, Johnson, and Bush, in review; Lenz,
Johnson, and Bush 2016) was implemented in a group of transtibial amputees. This work was
conducted under an approved Human Subject’s IRB protocol with informed consent (#14-089M).

47

3.3.1 Participants
Participants included eight transtibial amputees, one with bilateral amputation and seven
with unilateral amputation. All amputations occurred at least one year prior to the study. Reasons
for amputation included traumatic injury, diabetes, and infection. None of the participants had
any other disabilities or needed the use of an assistive device such as a walker or cane. At the
time of testing, participants were free of any sores on their residual limb.

3.3.2 Prosthetic Componentry
All participants had been seen within six months by a certified prosthetist who deemed
their current prosthesis was fitting well and had a proper alignment and fit. All participants were
using the same type of prosthetic suspension; a gel liner interface with pin/lock mechanism.

3.3.3 Test Procedure
The testing procedure consisted of two parts. First, to allow for measurement and capture
of the limb motion and anatomical landmarks, a duplicate socket was manufactured out of a clear
proprietary thermoplastic material commonly used in fabrication of prosthetic “test sockets” for

48

clinical use leading up to a definitive prosthetic socket (Figure 3-1 a, b). The duplicates were
developed to match the alignment and fit of the participant’s prosthetic socket.

Figure 3-1. Example of clear thermoplastic duplicated socket (a, b) for one participant compared
with their original opaque prosthesis. Componentry and alignment was maintained with only the
socket exchanged for the experimental test configuration (a, b). The duplicated socket allowed
the cameras to track the markers within the socket whereas the original socket does not.

The second portion of the study was a kinematic assessment. First a comparison of their
gait was made between the original device and the new duplicated clear socket. Kinematics data
were obtained so residual limb motions within and relative to the socket could be computed.
3.3.3.1 Comparison of Devices
To ensure the duplicated socket yielded identical walking patterns, a kinematic
assessment was completed with both the original and duplicated sockets. Kinematics were
obtained for the lower extremity and trunk joint angles during walking. From this, a Gait Deviation
Index was calculated for the walking trials (Schwartz and Rozumalski 2008). Secondly, the
participant’s duplicated prosthetic was assembled with their existing pylon, foot and suspension

49

components and data of lower extremity and trunk kinematics were captured with the same Gait
Deviation Index. The two data sets were compared to determine if significant changes were
introduced between prosthetic devices within a participant.
3.3.3.2 Measurement of Limb Motion within the Socket
The production of the clear sockets allowed reflective thin-disc motion capture markers
(9.5 mm diameter and 0.2 mm thick) to be positioned inside on the gel liner. Spherical markers
were also placed outside on the socket so that static anatomical locations and movement of the
limb relative to the socket could be calculated. Anatomical locations of interest included: anterior
tibial tuberosity, anterior tibial crest, distal tibial cut end, lateral proximal fibular head, distal fibular
cut end, intersegmental locations along the tibia/fibula and soft tissue medial limb proximal to
distal (Figure 3-2). All markers were positioned the same on all participants except two. On the
second participant, the limb contained a shorter fibula and only two markers were placed along
the lateral aspect of the limb, omitting the mid fibula. Participant eight had an even shorter and
in this case the mid fibula and mid tibial crest markers were omitted.

Figure 3-2. Calculation of displacements from anatomical marker locations on the gel liner within
the prosthetic socket

50

3.3.4 Analysis
A twelve camera motion capture system (Vicon Motion Systems Ltd.; Oxford, UK) tracked
the locations of all markers in three-dimensional space; the system was calibrated before each
data collection. Locations of markers were reconstructed using Vicon Nexus 2.3 to determine the
three-dimensional location of each individual marker within the global coordinate system.
Four parameters were developed for quantification of the residual limb within the
prosthetic socket: displacement, strain, displacement relative to the prosthetic socket, and rotation
of the limb relative to the prosthetic socket. With these data sets movements of the gel liner within
the socket and relative to the socket were calculated.
A displacement field was calculated based on how individual markers moved relative to
each other on the gel liner. Inter-marker distances were computed as the magnitude of the vector
between two markers. Inter-marker distance definitions on the limb relative to their anatomical
representation can be found in Figure 3-2.

Inter-marker distances were tracked during an

individual gait cycle and the displacement was defined as the maximum inter-marker distance
minus the minimum inter-marker distance during a single gait cycle. This was repeated for five
gait cycles per participant, averaged within participants, and also averaged across all participants.
Following inter-marker displacements, calculation of strain was completed. Strain was
defined by the change in length (i.e. the calculated displacement) divided by an original intermarker distance that was measured in a static standing trial of the participant.
To determine if the prosthetic device was moving relative to the residual limb, relative
socket displacement calculations were conducted. First, a local coordinate system (LCS) was
defined. The origin of the local coordinate system 𝑂𝐿𝐶𝑆 was defined as the midpoint between the
malleoli. The local coordinate system was created by a superior unit vector in the z direction
based on an axis passing from the distal end, local coordinate system origin, to the midpoint of
51

the knee joint center (Equation 1). This was achieved from medial and lateral markers on the
femoral epicondyles and the malleoli taken during the static standing trial.
𝑘̂ =

⃗ 𝐿𝐶𝑆
⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗
0.5∗(𝐾𝑛𝑒𝑒
𝐾𝑛𝑒𝑒𝑀𝑒𝑑 )− 𝑂
𝐿𝑎𝑡 + ⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗
⃗ 𝐿𝐶𝑆 |
⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗
⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗
|0.5∗(𝐾𝑛𝑒𝑒𝐿𝑎𝑡 + 𝐾𝑛𝑒𝑒𝑀𝑒𝑑 )− 𝑂

(1)

Next, a unit vector passing from the medial malleolus to the lateral malleolus was created, 𝑣̂
(Equation 2).

𝑣̂ =

⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗
⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗
(𝐴𝑛𝑘𝑙𝑒
𝐿𝑎𝑡 − 𝐴𝑛𝑘𝑙𝑒𝑀𝑒𝑑 )
⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗
|(𝐴𝑛𝑘𝑙𝑒
𝐴𝑛𝑘𝑙𝑒𝑀𝑒𝑑 )|
𝐿𝑎𝑡 − ⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗

(2)

Next an anterior unit vector was created from the cross product of 𝑘̂ and 𝑣̂ unit vectors. The third
lateral unit vector was created from the cross product of 𝑗̂ and 𝑘̂ unit vectors. Finally a rotation
matrix was created from unit vectors of the local coordinate system (Robertson et al. 2014).

𝑅𝑠ℎ𝑎𝑛𝑘

𝑖̂𝑥
= [ 𝑗̂𝑥
𝑘̂𝑥

𝑖̂𝑦
𝑗̂𝑦
̂
𝑘𝑦

𝑖̂𝑧
𝑗̂𝑧 ]
̂
𝑘𝑧

(3)

This rotation matrix was multiplied by x, y, z global coordinates of all markers to transform into
the local coordinate system (Equation 3).

Also, appropriate sign conventions were applied

depending upon direction of walking or whether it was a right or left residual limb.
3.3.4.1 Relative Motion
Following transformation into the local coordinate system, six locations were tracked
relative to the socket during walking. The locations corresponded to the tibial tuberosity, tibial cut
end, fibular head, fibular cut end, medial proximal soft tissue and medial distal soft tissue (Figure
3-2). The six marker locations were with respect to a marker placed on the prosthetic pylon to
track the relative movement between the device and the residual limb in the local z direction,

52

which was along the shank (Figure 3-3). In the local coordinate system, a vector was defined as
the difference in z direction components of one of the six markers within the socket minus the z
component of the prosthetic pylon marker. Throughout a single gait cycle the maximum distance
and minimum distances were located and the limb displacement relative to the prosthesis was
defined as the maximum minus minimum distance in the local z direction (Figure 3-3).

Figure 3-3. Analysis of vertical displacement and rotation relative to the prosthetic socket during
a gait cycle. ∆ Z symbolizing the change in marker location for relative displacement.

Rotation of the limb with respect to the socket was calculated for three regions on the limb
in the anterior, medial and lateral aspects. A vector was created between the two markers located
on the prosthetic pylon as a reference vector for the rigid body. Then, three independent vectors
were created in the anterior, medial and lateral regions of the residual limb by using the tibial
tuberosity and distal tibia markers for the anterior vector, followed by fibular head and distal fibula
for the lateral region vector, with lastly the two medial markers located proximally and distally for
53

the medial vector. Each of these vectors was used to calculate the angle between the limb vector
⃗
and the prosthetic pylon vector using Equation 4 where 𝐴 represents the pylon vector and 𝐵
represents one of the three region vectors of the limb. This was repeated for five gait cycles within
each participant and across all participants consistently.
⃗
𝐴𝑥𝐵

𝜃 = arctan ( 𝐴 ∙ 𝐵⃗ )

(4)

All reported measures were analyzed in five gait cycles per participant, reported as an
average with standard deviation within all individual limbs.

3.3.5 Statistical Method
Statistical comparisons were conducted within the four key parameters of displacement,
strain, relative strain and relative rotation. A one-way analysis of variance (ANOVA) followed by
Tukey’s post-hoc test was conducted to compare differences between the displacements and
separately to compare between the strains (MATLAB, 2012a). Statistical analysis using a oneway ANOVA with Tukey’s post-hoc was conducted on all nine limbs to compare displacements
and rotations relative to the prosthetic socket (MATLAB, 2012a).

3.4 Results
3.4.1 Participants
Mean participant age was 57.9 +/- 8.2 years and on average 14.5 years since amputation.
Table 3-1 provides details of participants. Two participants (2 and 8) had different shaped limbs
which did not allow for the same marker placement. The slightly changed marker set only
changed calculations for two measures: displacements and strains. Therefore, individual analysis
for these two patients were performed but were not included in the statistical analysis.

54

Table 3-1. Participant characteristics. Residual limb length defined as inferior edge of patella to
distal end of the stump. Mobility grade scale classifies an individual’s ability to ambulate or
navigate their environment. (Gailey et al. 2002). Level K3 is defined as the participant has the
ability or potential for ambulation with variable cadence - a typical community ambulator with the
ability to traverse most environmental barriers and may have vocational, therapeutic, or exercise
activity that demands prosthetic use beyond simple locomotion.

3.4.2 Displacements on the Residual Limb
Highest displacements in the residual limb analysis were observed at the distal fibula,
distal tibia and medial side (Table 3-2 a). These areas including the fibular head had significantly
more displacement compared with transverse direction displacements as assessed by the oneway ANOVA and Tukey’s post hoc for seven participant limbs with consistent displacements
measured (p < 0.05). On average the distal tibia, distal fibula and medial soft tissue regions
showed 5.83, 5.93 and 6.34 mm of displacement respectively.

3.4.3 Strains on the Residual Limb
Greatest strains occurred in the proximal tibia, distal tibia and distal fibular regions (Table
3-2 b). These were statistically significant differences when compared to the transverse direction
strains by statistical analysis of a one-way ANOVA followed by Tukey’s post hoc (p < 0.05). Distal
tibia and distal fibula region values were on average 11% strain followed by 10% strain in the
proximal tibia region, with a lowest strain of 6% occurring in the proximal fibula region.
55

Table 3-2. (a) Displacements for participants. Reported as average of five gait cycles plus or
minus standard deviation. (b) Strains for participants. Reported as average of five gait cycles
plus or minus standard deviation.

3.4.4 Displacements Relative to the Prosthetic Socket
Lateral and medial anatomical landmarks had the largest displacements relative to the
prosthetic socket. On average, the greatest relative displacement of 30.7 Âą 11.4 mm was
observed in the medial proximal soft tissue region, followed by the fibular head with 27.3 Âą 10.7
mm. In comparison, anterior displacements of 11.0 Âą6.3 to 13.3 Âą 6.4 mm at the tibial tuberosity
and distal tibial prominence exhibited significantly less displacement relative to the prosthetic
socket. Furthermore, the distal fibula displaced statistically less than the fibular head and medial
proximal soft tissue regions (p < 0.05). Additionally, “gapping”, or a space between the socket

56

and residual limb was observed visually in six participants and can be seen for one participant in
Figure 3-4.

Figure 3-4. Demonstration of distal gap between residual limb and prosthetic socket within the
socket just prior to initial contact and loading during walking

3.4.5 Angular Rotation of Limb Relative to the Socket
Angular rotation along the anterior, lateral and medial regions of the residual limb relative
to the prosthetic socket showed rotations ranging from 0.8 to 10.8 degrees across all participants.
On average the anterior, lateral and medial rotations were 5.3 Âą 3.6, 5.2 Âą 3.5, and 5.4 Âą 3.9
degrees respectively. These data did not yield statistically significant differences when compared
using a one-way ANOVA across all participants. Overall all regions uniformly rotated relative to
the socket as measured as a change in rotational displacement during a gait cycle.

57

3.5 Discussion
The goal of our study was to investigate residual limb displacements within the socket,
and displacements and rotations relative to the prosthetic socket in transtibial amputees during
walking. Results overall confirmed that higher displacements and strains were observed distally
as compared to proximally. The reason for this increased distal displacement and strain is related
to the suspension method. At the base of the gel liner, a pin system locks to the prosthetic socket,
leg and foot.

This causes elongation of the gel liner due to the suspended weight.

Our

displacement data are the most comprehensive data set currently available; we have computed
displacements at six regions in the longitudinal direction, two in the transverse direction, three
regions of rotation of the prosthetic relative to the leg and relative displacement of the prosthetic
to the leg.
Gholizadeh et al. utilized motion capture to assess vertical displacement of the limb
relative to the socket along the lateral aspect of the residuum (Gholizadeh et al. 2012). However
this was not conducted during walking but rather a progression of full-weight bearing, semi-weight
bearing, and non-weight bearing. Gholizadeh et al. reported 0-6 mm change in displacements
which are 25 mm lower than our study. The controlled loadings in Gholizadeh’s study are believed
to result in the lower displacement values because the limb did not experience inertial dynamics
typical during the gait cycle.
Our data indicates that highest displacements occurred in the regions of the distal fibula,
distal tibia and medial soft tissue. These regions are all close to the insertion point of the pin to
the locking mechanism. Thus, the entire weight of the prosthetic device is suspended from the
pin at the distal end of the gel liner and is the cause of these larger distal displacements.
Assuming the gel liner is moving with the skin surface, these high gel liner displacements are
would also cause high displacements of the skin. In some cases, it was clinically observed that

58

a clear separation occurred between the distal limb and the prosthetic socket. Knowing the gel
liner displaces, there are two possible interactions that can occur between the liner and the skin:
1) no slip occurs or 2) the liner slips with respect to the skin. If it is assumed that the gel liner is
moving with the skin surface, these high gel liner displacements would also cause high
displacements of the skin. If the gel liner is, however, slipping with respect to the skin, a suction
hematoma at the distal residual limb may be creating erythemas or dilatation of the blood
capillaries (Levy 1995; Levy & Barnes 1956). Both cases of stretch, or slip of the liner on the skin
will lead to increased forces on or within the tissue which have been shown to reduce regional
blood flow and lead to conditions that produce ulcers (Manorama et al. 2010; Manorama et al.
2013).
Displacements were converted to strains in the study. Both displacement and strain data
create a sense of where elongation of the liner occurs. Sanders et al. documented different
responses of liners in tensile test experimental setups and results were reported in kPa, noting
significant differences between gels, elastomers and urethane based liners (Sanders et al. 2004).
While many variables were constrained in our recruitment of participants, gel liners were not
controlled for, but all were documented. One previous study reported maximum relative strain of
proximally located markers to be about 8-10% in participants during stepping down activities
which is similar to what we found (Papaioannou et al. 2010). Prosthetists should consider liner
material properties when determining selection for patients experiencing high strains distally.
Clearly, significant movement occurs distally with the gel liner. Clinically, this is important to note.
Although a stiffer liner could be used to decrease displacement/strain, it may cause other issues.
For example a stiffer liner could result in higher loads over bony prominences because it is less
forgiving.
One limitation of this study is that only one type of prosthetic suspension was studied. To
achieve pin locking suspension, a gel liner interface is required. Therefore, another limitation is
59

that residual limb movements described are always with respect to gel liner motion because
measurement was taken from the surface of the gel liner, not the skin surface itself. We are
unable to state definitively what the underlying tissue is doing with respect to the gel liner. Two
possible assumptions are that 1) the liner moves and stretches with the skin and 2) the liner slides
along the skin. Both of these scenarios will be studied computationally in future work.
Limb displacement relative to the prosthetic socket revealed less displacement on the
proximal and distal anterior tibial locations as compared to medial and lateral locations. This
statistically significant finding indicates the socket is constraining anteriorly which may be related
to the fact that we are evaluating gait which is a sagittal plane motion with loading from the
prosthetic onto the leg primarily in the anterior/posterior regions which would reduce shifting of
the device. The lateral sides of the residual limb/ prosthetic device are less constrained during
gait.

High displacements in the proximal fibula region are clinically concerning because

prosthetists often create reliefs for the fibular head when designing a modified PTB style socket.
The clinical thought is to allow this bony prominence some extra room for comfort, however with
on average 27.3 mm of vertical displacement relative to the prosthetic socket, it is likely the fibular
head prominence is displacing outside of the designated relief and potentially contributing to this
regions common formation of ulcerations. We hypothesized higher displacements and the data
provided a more detailed view of limb displacement than had been previously documented but
within the documented ranges in the literature. Displacement during swing relative to stance
phases of gait, proximal displacements averaged 41.7 mm across multiple gait cycles for a single
transtibial amputee (Sanders et al. 2006). Another study compared two liners demonstrating the
Seal-In X5 liner decreased vertical pistoning by 71% compared to the Iceross Dermo liner
(Gholizadeh et al. 2012). No statistically significant differences in residual limb rotation throughout
the gait cycle, confirms our hypothesis that the within socket motion is primarily a vertical pistoning
motion rather than twisting motion of the limb relative to the socket. This study presents a larger

60

sample size with more refined detail than previously existed to our knowledge, positively
contributing to the knowledge of transtibial limb motion within a prosthetic socket with the
utilization of pin/lock mechanism suspension.

These data provide a comprehensive set of

information on the interface displacements and can be implemented clinically to design systems
with reduced pistoning which in turn has the potential to decrease skin wounds.

3.6 Acknowledgements
The authors would like to thank Patrick Logan, CPO for his support and encouragement of the
research. Additionally, Sarah Veldman, Kate Clarke and Curtis Ackermann are thanked for their
assistance in data extraction and marker tracking.

3.7 Funding
Michigan State University Graduate Office Fellowship and Michigan State University Mechanical
Engineering Fellowship

61

4. INSTRUMENTED TRANSTIBIAL SOCKET FOR EVALUATING SHEAR AND
NORMAL FORCE: A CASE STUDY

62

4.1 Introduction
Pressure ulcers are areas of tissue damage caused by loading on the skin from various
body-interface conditions. Ulcers can be either superficial or deep in nature depending on the
loading conditions (Mak et al. 2010). Within an amputee’s transtibial socket, controlling the friction
is essential for minimizing shear loads on the skin which could lead to the formation of pressure
ulcers. The socket interface with the residual limb ranges from direct skin contact to gel liner
interfaces. A gel liner is typically made of urethane, silicone or thermoplastic elastomer and is
used to provide cushioning with respect to the rigid socket (Spires et al. 2014). Gel liners are
most commonly prescribed for persons with below knee amputations (Boutwell et al. 2012). When
a patient’s limb fluctuates in volume throughout the day, additional sock layers are prescribed to
be worn to fill extra space within the socket for a more comfortable fit. However, additional layers
are not always necessary and could increase interface loads if worn when not indicated.
Experimental investigation of loading within the prosthetic socket is needed in these conditions to
understand the loading mechanics occurring at this interface.
Studies have documented pressure distributions using pressure sensor
technology within the socket and to compare different socket types (Pearson 1974; Meier
1973; Leavitt et al. 1970; Appoldt et al. 1969; Naeff & van Pijkeren 1980; van Pijkeren et
al. 1980). Most recently, pressure distribution at the socket to limb interface has been
used to compare differences in prosthetic componentry, suspension types, and liners (Beil
et al. 2002; Hossein Gholizadeh et al. 2014; Ali, Abu Osman, et al. 2012; Wolf et al. 2009;
Boutwell et al. 2012; Eshraghi et al. 2013; Gholizadeh et al. 2015; Ali, Osman, et al. 2012).
These studies only addressed pressure which is related to the normal loading. Shear
forces have been shown to be more detrimental to skin health (Manorama et al. 2010;
Manorama et al. 2013), however, none of these studies addressed shear and normal
forces at the socket to limb interface.
63

Load cells have been used in prosthetic research to understand moments at the interface
between the pylon and the base of the prosthetic socket. Neumann et al. instrumented a load
cell at the base of the prosthetic socket to obtain moment data for below knee amputees walking
on various terrains and curved pathways to measure real world situations (Neumann et al. 2013).
In one study, transducers were also used to measure shear stresses in two orthogonal directions
on a plane flush with the inside of the socket (Sanders et al. 1992). Three participants were tested
and each socket was lined with a custom molded foam insert, designed to fit without the use of
an additional sock or nylon sheath. This allowed for shear stresses to be measured directly at
the skin. To further the understanding of shear stresses at the surface of the residual limb,
Sanders et al. improved on the original work expanding shear stress measurement to thirteen
locations on two patients with transtibial amputation (Sanders et al. 1997). In the type of
suspension method used by Sanders et al., a gel liner interface was not required and the effect
of multiple liners was not addressed.
Most recently Schiff et al. instrumented load cells into a transtibial socket utilizing sleeve
suspension to further explore shear forces for amputees with and without distal tibia-fibular bone
bridges (Schiff et al. 2014). However, sleeve suspension systems function mechanically different
than gel interfaced pin suspension by gripping the prosthetic device from the outside of the
prosthetic socket rather than at a single distal pin location. Additionally, pin suspension systems
with gel liner interface are the most common attachment for below knees amputees and as such
warrant investigation (H. Gholizadeh et al. 2014). Extensive shear force information for prosthetic
users is currently lacking, more experimental work is needed to understand interface mechanics
to improve our understanding of ulcer formation (Schiff et al. 2014). Force data gathered
experimentally also has the potential to be used an input to a finite element modeling so deep
tissue stresses can be obtained and multiple conditions studied.

64

The objective of this research was to instrument a single transtibial prosthetic socket with
a two axis load cell to measure longitudinal shear force and normal force during walking. Force
data located at the mid fibular region of the residual limb was desired as input for a finite element
model and to further investigate interface mechanics of a gel liner interface. Experimental data
comparisons were also evaluated with a gel liner, additional sock ply and removal of gel liner to
allow contact with the skin. It was hypothesized that normal and shear forces would increase with
increasing sock ply due to the added thickness within the socket allowable volume.

4.2 Methods
One male right unilateral transtibial amputee, 1.77 m in height, mass of 85.5 kg and
mobility grade of K3 (Gailey et al. 2002), participated in this study. The participant provided written
consent to complete the study. He underwent amputation twenty-nine years prior due to a
traumatic accident.

His current prosthetic device componentry included a 9 mm Alpha

Willowwood gel interface liner with pin locking suspension.

At time of testing, no pain or

discomfort with his prosthetic device was reported. He had been seen within a month by a
certified prosthetist who deemed his socket interface to be appropriate and alignment of his
prosthesis to be acceptable.
First, to allow for force measurement, a duplicate socket was manufactured out of a clear
thermoplastic material commonly used in fabrication of prosthetic “test sockets” for clinical use
leading up to a definitive prosthetic socket. The duplicate was developed to match the alignment
and fit of the participant’s prosthetic socket. A duplicated socket was required for the insertion of
a load cell within the inner socket surface. A hole was cut in the wall of the prosthetic socket in
order to mount the load cell flush within the inner surface which contacted the limb.

The

participant’s original socket would have been compromised if used for the experimental
procedure.

65

Prior to load cell instrumentation a kinematic assessment was completed with both the
original and duplicated sockets to ensure the duplicated socket yielded identical walking patterns.
Kinematics were obtained for the lower extremity and trunk joint angles during walking. From
this, a Gait Deviation Index was calculated for the walking trials (Schwartz & Rozumalski 2008).
Secondly, the participant’s duplicated prosthetic was assembled with their existing pylon, foot and
suspension components and data of lower extremity and trunk kinematics were captured with the
same Gait Deviation Index. The two data sets were compared to determine if significant changes
were introduced between prosthetic devices within a participant.
To install the multi-axis load cell, a 4 cm diameter hole was created in the duplicated
prosthetic socket. The hole’s center was located 8.5 cm below the patient’s fibular head on the
lateral aspect of the residual limb. This location was selected because of the presence of muscle,
complete contact with the load cell and avoidance of a bony prominence. The absence of contact
would not allow for force measurement with the mounted load cell.
An OptoForce HEX-70-CE load cell (OptoForce; Budapest, Hungry) was mounted
securely to the prosthetic socket as seen in Figure 4-1. The removed material from the prosthetic
socket wall was mounted to the recording surface of the load cell with counter sunk screws.
Mounting the removed piece to the load cell allowed for a consistent coefficient of friction across
the entire socket. Additionally, it maintained the original curvature of the internal socket surface
therefore preserving the clinically developed design of the socket. Shear load was measured
along the long axis of the socket. Normal force was perpendicular to the internal curvature of the
socket and designated as the z direction of force.

66

Figure 4-1. a) Top view of instrumented socket demonstrating flush internal curvature. b)
Oblique view of load cell securely affixed from the outside of the prosthetic socket.

The instrumented prosthetic was assembled with the patient’s every day original componentry
from below the socket to pylon interface, maintaining the alignment of the device as a whole
(Figure 4-2). The patient walked with the instrumented socket for an adjustment period.

Figure 4-2. Assembled instrumented prosthesis as worn by the participant.

67

Three walking conditions were conducted while two axes of force data were collected from
within the socket. First, the participant wore his original gel liner within the socket, which was a 9
mm Alpha Willowwood gel liner (E = 50050 Pa). Secondly, a 3 ply sock was added over the gel
liner. Lastly, a hole was cut in the 9 mm gel liner, exposing the patient’s skin which allowed direct
force measurement at the skin surface during walking (Figure 4-3). The hole was slightly larger
than the recording surface on the load cell to ensure only skin contact was being recorded. The
skin protruded through the hole to the outer region of the liner and the mounting of the load cell
was not changed. For all three testing conditions, multiple self-selected walking trials were
collected from which eight gait cycles were analyzed in each condition. Data were collected from
the load cell at a 100 Hz sampling frequency.

Figure 4-3. Whole in gel liner to allow for contact and force measurement at the skin surface.

Force data were analyzed in gait cycles with respect to heel contacts and toe offs defined
as gait events for the residual and sound limbs.

Motion capture and force systems were

synchronized with an external trigger so that force data could be split into gait cycles based on
motion capture gait events. This analysis was performed using a custom Matlab code based on

68

motion capture data for landmarks defined on the bilateral calcanei and second metatarsal heads
(Miller 2009).

Within each condition, shear and normal force data from eight gait cycles were

normalized with respect to the gait cycle and averaged together. Force data during the gait cycle
were selected for each condition corresponding to the peaks during initial contact, early stance,
single support and swing. Based on the circular four centimeter diameter contact area, stresses
were also calculated for the three conditions.
Four statistical comparisons were conducted using an ANOVA followed by a Tukey’s
posthoc for each analysis (MATLAB, 2016a). Three way comparisons analyzed the experimental
normal and shear differences in force values across the three conditions including 1) gel liner
alone 2) three ply sock worn over the gel liner, and 3) a hole cut in the gel liner exposing the skin.
Separate peak force statistical analyses were conducted for these three comparisons within the
four regions of the gait cycle defined earlier: initial contact, early stance, single support and swing.
Level of significance was set to a p value of p < 0.05.

4.3 Results
The results obtained from experimental data collection using the instrumented socket
revealed differences in force throughout the gait cycle across condition types (Figure 4-4). First,
shear force data averaged across eight gait cycles revealed statistically significant differences
across all comparisons which included four ANOVA results across the three conditions (gel liner
alone, 3 ply sock with gel liner and the hole in the liner) with respect to the peaks identified during
the gait cycle (Table 4-1). During all four analyzed regions (initial contact, early stance, single
support and swing) of the gait cycle, shear force was greatest for the condition with the addition
of a 3 ply sock and least in the hole in liner condition with the gel liner alone falling in between (p
< 0.05).

69

Comparing normal forces, there were only two comparisons which were not found
statistically significant as a result of the ANOVA analysis within periods of the gait cycle (Table 42). First, during initial contact, the gel liner and hole in liner conditions were not statistically
different from each other but the three ply sock was significantly different from both. Secondly
during early stance, the three ply sock did not demonstrate a statistical increase in normal force
when compared with the gel liner alone condition, however later in single support all conditions
were statistically different from each other for normal force (p < 0.05). It is also notable that a
residual normal force during swing phase of gait were highest with the 3 ply sock addition, which
was statistically significant when compared with the gel liner and hole in liner conditions (p < 0.05).

Figure 4-4. Three conditions of normal and shear force during walking from heel contact to heel
contact on the right residual limb. Sashed lines in the same condition color represent the Âą
standard deviation of force data as it was analyzed across 8 gait cycles.

70

Table 4-1. Force comparison across three conditions for the four particular periods during the
gait cycle.

Gel Liner
3 Ply Sock
Hole in Liner

Fx Longitudinal Shear Force (N) (Avg Âą SD)
Initial contact Early Stance Peak Single Support Peak
0.79 Âą 0.33
7.68 Âą 0.16
7.01 Âą 0.25
2.86 Âą 0.25
10.01 Âą 0.21
9.85 Âą 0.24
0.39 Âą 0.13
3.17 Âą 0.49
4.29 Âą 0.38

Swing
0.97 Âą 0.22
2.96 Âą 0.21
0.40 Âą 0.23

Gel Liner
3 Ply Sock
Hole in Liner

Fz Normal Force (N) (Avg Âą SD)
Early Stance Peak Single Support Peak
77.53 Âą 2.15
57.06 Âą 0.79
78.21 Âą 3.06
61.64 Âą 1.17
56.39 Âą 3.36
49.49 Âą 0.90

Swing
5.24 Âą 0.88
11.52 Âą 1.39
4.32 Âą 0.68

Initial contact
4.99 Âą 1.52
10.84 Âą 1.39
4.71 Âą 1.45

Table 4-2. Initial contact and early stance statistical p values noting statistical differences
reported for normal force comparisons.

P Value

Initial Contact Fz Normal Force Comparison
Gel Liner vs 3 Ply Sock Gel Liner vs Hole in Liner 3 Ply Sock vs Hole in Liner
p < 0.05
p > 0.05
p < 0.05

P Value

Early Stance Fz Normal Force Comparison
Gel Liner vs 3 Ply Sock Gel Liner vs Hole in Liner 3 Ply Sock vs Hole in Liner
p > 0.05
p < 0.05
p < 0.05

Shear and normal stress during walking were calculated based on the contact area of the
load cell that was imbedded within the prosthetic socket. Shear stresses ranged from 0.4 – 7.66
kPa and normal stresses ranged from 2.7 – 61.9 kPa when evaluated across the three conditions
and throughout the gait cycle. For the gel liner condition, peak normal stresses were calculated
to be on average 61.7 Âą 1.7 kPa. On average during the entire gait cycle, normal stresses for the
gel liner condition were calculated to be 27.45 kPa. These calculated normal stresses were
compared to published stresses and distributed pressures.

71

4.4 Discussion

The goal of this study was to evaluate how normal and shear forces during walking for a
transtibial amputee. Additionally, force data were obtained for three specific conditions 1) gel liner
alone 2) three ply sock worn over the gel liner, and 3) a hole cut in the gel liner exposing the skin.
The inclusion of shear forces in experimental work is critical. Shear force has been proven
to contribute to increased blood occlusion when compared with normal force loading alone
(Manorama et al. 2010; Manorama et al. 2013) which in turn has been shown to lead to increased
tissue necrosis (Oomens et al. 2016). In an amputee population this is an important consideration
when evaluating limb interface mechanisms and the high risk for pressure ulcer formation on the
residual limb. Our results demonstrated that while normal forces throughout the gait cycle did not
vary greatly with the addition of a three ply sock, the shear forces were all significantly increased
with the added sock thickness.

However, we have measured a significant increase in shear

forces with increased sock thickness which may lead to increased blood occlusion and resulting
decreased nourishment of limb tissues. Wearing an extra sock thickness when not clinically
indicated may cause limb irritation and conditions related to ulcer formation due to increased
shear forces at the interface. Therefore, patients should be educated on sock usage, and
clinicians should consider the proper indications for prescribing an additional sock ply.
For the condition with the hole in the gel liner, shear and normal force measurements were
collected at the surface of the skin. A challenge of conducting experimental research in below
knee amputees wearing a gel liner is the inability to directly access the surface of the skin for data
measurement. Historically, this is a reason why thin film pressure measurement devices were
used extensively in published literature because they are able to be placed beneath the surface
of the gel liner, while still utilizing the typical suspension method. However, shear forces are not
able to be measured with pressure sensors. Our results indicated significant decreased in shear
72

and normal forces throughout the gait cycle when compared to the intact gel liner condition. A
possible reason for the decreased forces may be the absence of material thickness, therefore
allowing the tissue less bounded constrain and greater regions to displace within the socket.
Stresses were calculated based on the contact area of the load cell within the socket. For
the gel liner condition, peak normal stresses were calculated to be on average 61.7 Âą 1.7 kPa.
Sanders et al. also measured normal stresses in the mid fibular region to be on average 61.3 kPa
(Sanders et al. 1997). Additionally, average normal stresses for the gel liner were computed to
be 27.45 kPa throughout the entire gait cycle. Gholizadeh et al. conducted an experimental study
with pressure sensors mounted on transtibial limbs and found an average pressure throughout
the gait cycle to be 36.05 Âą 11.4 kPa in the mid to distal fibular region (Hossein Gholizadeh et al.
2014). Considering large standard deviations across the pressure measurement in Gholizadeh’s
study, our single case study falls within their reported values. However, limited shear force and
stress data are available in the below knee amputee population. Schiff et al. evaluated correlation
coefficient comparisons of stresses through the use of load cells instrumented in the socket wall
but magnitudes were never presented (Schiff et al. 2014).

Our case study provides new

information, particularly with regard to shear forces in addition to normal forces within the
prosthetic socket.
A limitation of this study includes the singular location in which force was measured. The
targeted region was selected based on tissue thickness in this anatomical region of the residual
limb.

Future studies could also include more measurement locations to compare regional

anatomical variances in soft tissue versus bony prominences. Secondly, a limitation of this
methodology includes possible changes introduced due to the presence of a hole in the gel liner
such as elongation changes to the gel liner or irritation caused between the skin/rigid socket wall
interface.

73

Overall, this work contributes significantly to the understanding of shear forces at the
socket to limb interface. Many patients wear different interface thicknesses during the day and
this case study highlights a single sample size of how these forces change. The addition of shear
force measurement in addition to normal force is important for considering the underlying tissue
influences for blood perfusion and tissue health when shear is present. Finally, these results are
also helpful for creating finite element models which can be based on experimentally captured
data for boundary conditions and loading inputs.

74

5. FINITE ELEMENT ANALYSIS OF THE SOCKET TO LIMB INTERFACE WITH
EXPERIMENTAL DATA INPUTS

75

5.1 Introduction
The mechanical interface between a patient’s residual limb and their prosthetic device is
a complex loading condition. When a prosthesis is not fitting well, it often leads to the formation
of ulcers.

While experimental and modeling research has been conducted in this patient

population, there is much yet to be understood about key contributors to ulcer formation. Finite
element (FE) can be used to study the effects of different prosthetic liners on superficial and deep
tissue stresses.
FE has been utilized in understanding interface biomechanics, development of sockets
with computer aided design (CAD)/ computer aided modeling (CAM) and approximation of deep
tissue stress (Zhang et al. 1998). The model is dependent upon material properties, geometric
data, loading characteristics, boundary conditions and initial conditions. While a FE model may
result in a solution, it does not guarantee the model’s value or accurate representation of the
physical system. Therefore, combining a FE model with experimental data helps insure realistic
outputs.
Early modeling in prosthetics simplified the geometry of the residual limbs and the
prosthesis design, often constraining the model with a static ground reaction force at heel contact.
Also, lack of experimental data at the socket to limb interface caused researchers to make many
assumptions on constraints and external loading. Models did not represent the complexity of the
bone structures within the limb, nor did they model any of the biological tissues as non-linear
materials (Brennan & Childress 1991; Reynolds & Lord 1992; Sanders & Daly 1993; Steege &
Childress 1988; Silver-Thorn 1991; Quesada & Skinner 1991; Zhang et al. 1995). A summary of
initial prosthetic modelling literature is provided in Appendix A.
As imaging techniques improved so did the complexity of models.

Biplanar x-ray,

magnetic resonance imaging (MRI), CAD and computed tomography (CT) were used to create

76

realistic geometries for modeling (Lee et al. 2004; Zhang & Roberts 2000; Lee & Zhang 2007;
Zachariah & Sanders 2000). However, a limitation of these studies was a continued assumption
that soft tissues could be modeled as linearly elastic materials. A summary of models using more
advanced geometry, with linear elastic materials is located in Appendix B.

5.1.1 Deep Tissue Finite Element Analysis (FEA) in Amputees
More recent models have multiple tissue layers, interactions between those layers and
mathematical constitutive models of non-linear material properties to more closely mimic
biological tissues. The development of these improved models then allowed for the investigation
of deep tissue injury (Portnoy et al. 2008; Portnoy, Siev-Ner, Yizhar, et al. 2009; Portnoy, SievNer, Shabshin, et al. 2009; Portnoy et al. 2011; Portnoy et al. 2010; Gefen et al. 2008). Key
findings for transtibial amputees indicate higher stresses accumulating at the tissue to bone
interface rather than more superficial tissue layers. However, the use of experimental data inputs,
as boundary conditions is still lacking.

5.1.2 Recent FEA Using FEBio
Finite Elements for Biomechanics (FEBio) is a software platform customized for complex
questions in biomechanics with numerous constitutive models available in the package, and is
open source (Maas et al. 2012; Meng et al. 2013; Galbusera et al. 2014; Perduta & Putanowicz
2015). Recent research is limited with regard to use of non-linear, neo-Hookean materials
implemented in FEBio. The available work reported for other body regions included heel ulcers,
buttock tissue damage in spinal cord injury, and a single transtibial patient was summarized in
Appendix C (Sengeh et al. 2016; Levy et al. 2015; Shoham et al. 2015). Specifically, Levy et al.,
investigated tissue stresses in heel ulcers where they implemented non-linear, neo-Hookean
materials for the skin layer with shear and bulk moduli consisting of 31.9 and 3179.37 kPa (Table
5-1) (Levy et al. 2015). Shoham et al., also implemented neo-Hookean materials in their model
77

for skin and muscle with the same skin values as Levy et al. and muscle values as seen in Table
5-1 (Shoham et al. 2015). For our work, FEBio was used to model deep tissue stresses in a
region of a transitibal limb with experimental data supporting boundary inputs.
Table 5-1. Mechanical properties of the model components selected for the FE models.

5.1.3 Current Need for FEA in Prosthetics
Particularly relevant to prosthetics, FEA can be used to understand underlying tissue
stresses which can lead to ulcer formation due to different prosthetic gel liners and interface
conditions. Additionally, inputs to the FE models can utilize experimental data captured within
the prosthetic socket, such as forces and displacements. Doing so would contribute new analyses
to the literature which provide realistic data sets based on measured boundary conditions. For
example, previous research based on experiments has shown increased shear forces contribute
to increased blood occlusion and consequently increased risk for ulcer formation (Manorama et
al. 2010; Manorama et al. 2013). Based on the findings by Manorama et al., it is desirable to
identify how various gel liner properties influence deep tissue mechanics under shear and normal
loading conditions. Thus, the objective of this work was to determine the stresses within
the skin and muscle for a transtibial amputee limb, constrained with experimental
conditions of displacement and force data. The goal was achieved by creating 29 models,

78

beginning with simplified analyses and advancing to multi-layer models representing various gel
liners with underlying skin, muscle and bone.

5.2 Methods
During the development of the models, simplified analyses were first implemented to
investigate non-linear responses of neo-Hookean materials. Next, complex models consisting of
four layers were developed with the gel liner, skin, muscle and bone. Then, the contact interface
between the skin and gel liner was varied to evaluate under which conditions the gel liner would
slip relative to the skin. Understanding the slip and no slip conditions and their effects on skin
Von Mises and shear stress provided an improved understanding how various interface conditions
may lead to ulcer formation.

5.2.1 Simplified One and Two Layer Models
The purpose of the simplified models were to investigate responses of neo-Hookean
materials to various loading conditions. Simple cubes of neo-Hookean material were created to
evaluate the influence of separate compressive pressure and prescribed displacements on this
quasi-incompressible, hyperelastic material and compare to theoretical results (Figure 5-1a). First
the entire cube of neo-Hookean material was defined with material properties for Young’s
modulus (E = 50 kPa) and Poisson’s ratio (𝑣 = 0.49). Two separate models were executed with
two different distributed pressures, 10 Pa and 5 kPa respectively, in which normal stress, strain
and displacement were analyzed. Based on experimental force data collected in Chapter 4, five
kPa was selected as the loading condition. The 10 Pa load was arbitrarily selected to perform an
analysis in which the neo-Hookean material responded in the linear elastic range. Next, a 6 mm
prescribed displacement was applied parallel to the top surface of the cube in the positive x
direction.

79

Figure 5-1. a) Square neo-Hookean model b) Square neo-Hookean model with each layer
divided evenly in height c) Square neo-Hookean model with uneven layers

A simple two layer neo-Hookean model was also created to investigate the layering
method using tied interfaces between layers (Figure 5-1b and 5-1c). Layers were first model as
even thicknesses and secondly where the top layer was 0.2 m thick and the bottom layer was 0.8
m thick. Initially, the two layers were modeled with the same material properties (E = 50 kPa and
𝑣 = 0.49). Next, the top layer was varied slightly with a material that had a reduced stiffness (E =
40 kPa and 𝑣 = 0.49), while maintaining the bottom layer properties (E = 50 kPa and 𝑣 = 0.49).
All two layer model variations were loaded with a 5 kPa compressive distributed pressure with the
bottom XY surface of the material constrained for all translations and rotations.

5.2.2 Realistic Four Layer Models
The inclusion of the four layer models was to create a realistic layered region of a
transtibial amputee’s limb consisting of the gel liner, skin, muscle and bone. Once the model was
established, focus was given to the investigation of various gel liner material properties and
contact interactions on underlying tissue stresses.

80

5.2.2.1 Mechanical Properties
Mechanical properties of bone, muscle, skin and various gel liners were adapted from the
literature (Sanders et al. 2004; Levy et al. 2015; Shoham et al. 2015) (Table 5-1). Specifically,
the bone was assumed to be a rigid body. The muscle and skin were nearly incompressible and
hyperelastic, represented by a neo-Hookean material model with a strain energy density function
W (Equation 1):
𝑊=

𝐺𝑖𝑛𝑠
2

1

(𝜆12 + 𝜆22 + 𝜆23 ) + 𝐾(ln 𝐽)2
2

(1)

Where, 𝐺𝑖𝑛𝑠 represents the shear modulus, 𝜆1 , 𝜆2 & 𝜆3 are principal stretch ratios, 𝐾 is the bulk
modulus and 𝐽 = det(F) , where F is the deformation gradient tensor. This material type was
selected because it has non-linear stress-strain behavior but uses a standard displacement-based
element formulation.
5.2.2.2 Modeling Region of Interest and Geometry for Realistic Model
First a rectangular geometry model of 100 mm x 100 mm was developed with layers of
bone, muscle, skin and a gel liner (Figure 5-2). Layer thicknesses were based on literature, 10
mm of bone, 20 mm of muscle, 3 mm of skin, and 9 mm of gel liner (Sanders et al. 2004; Zollner
et al. 2013; Tepole et al. 2011; Levy et al. 2015; Shoham et al. 2015). It should be noted that
muscle thickness across the entire fibula is uneven distally versus proximally.

The model

represents anatomy in the distal third of the fibula where muscle thickness is more consistent
(Figure 5-3).

81

Figure 5-2. a) 3D geometry of four layer model b) Simplified geometry of FE model containing
bone, muscle, skin and gel liner constrained by experimental displacements and fixed at the
boundary of the bone. The gel liner to skin interface was either modeled as a tied interface or a
frictional sliding interface depending on the analysis to demonstrate if slip occurred or not. A
tied interface between muscle/skin used a connection of two non-conforming meshes with a
high penalty factor to prevent modeled tissue separation. A rigid interface between bone/muscle
was also defined to not allow separation.

Figure 5-3. Lower extremity model of bones and generalized soft tissue focused on an
amputee’s anatomy of limited limb length. The anatomical area enclosed by the rectangle
represents the fibular region of interest for which tissue thicknesses were modeled based on
average tissue thickness in this region.

To examine robustness of the model with respect to geometry, additional model designs
were implemented. A cylindrical model was developed as an axis-symmetric comparison with
82

respect to the rectangular four layer model. Layer thicknesses for bone, muscle, skin and gel
were identical to that of the rectangular model with a cylindrical radius of 56.42 mm for all the
layer cross sectional areas resulting in an equivalent layer area of 0.01 m2.
5.2.2.3 Loading and Boundary Conditions for Realistic Model
Based on measured force data found experimentally during walking, a 5 kPa distributed
pressure was applied to the gel liner surface. An average regional displacement for the mid-fibula
of 6 mm (obtained from our experimental data) was used for a prescribed displacement and
applied at the surface of the gel liner (Figure 5-2).

5.2.2.4 Layer Contact Definitions for Realistic Model
Within the model tissue layers, the skin to muscle contact was treated as a tied interface
with high penalty factor to prevent the formation of gaps between the layers. In FEBio, a tied
interface can be used to connect two non-conforming meshes and it is assumed that the nodes
on both mated surfaces are connected. The muscle to bone interface was treated as a rigid
interface. The bone layer was constrained in all directions of translation and rotation as a rigid
body on the bottom surface. At the skin to gel liner interface, two model simulations were
completed to represent two clinical situations; 1) the gel liner slipped along the skin 2) no slip
occured with respect to the skin during loading (Figure 5-4). For the slip condition, a frictional
interface coefficient was set to a 0.3 based on Sanders et al. 2004. This allowed for slip to occur
if the coefficient of friction was not great enough to prevent tangential movement between the skin
and gel liner. For the no slip condition, a tied interface was modeled with the penalty factor set to
prevent layer separation.
The models with the condition of slip between the liner and skin interface utilized a sliding
with gaps contact. The slip conditions required xz face vertical constraints to properly define the

83

model. Although the vertical constrains were not necessary for the tied interface on the no slip
model, an analysis was conducted including xz face vertical limitations on displacement. The
purpose of running two simulations, one with and one without vertical constraints on the no slip
model, was to evaluate the influence of vertical constraints on tissue layer stresses.

Figure 5-4: a) No slip model and b) Slip model demonstrating contact differences between the
skin and gel liner interaction. As can be seen on the right, there is x direction translation
between the gel liner and skin.

5.2.2.5 Liner
Four models were created, each represented a specific liner type. The selected liners
were used by our participants, which included an Ohio Willowwood Alpha Liner, Ossur Iceross
Comfort, Ossur DERMO-6, and TEC Pro 18 (Table 5-1). Our experimental strain of the liner
during walking from Chapter 3 reported 6-11% strain which falls within this linear response region
(Sanders et al. 2004). Thus, each gel liner was modeled as a linear isotropic material based on

84

linear responses of these gel liners during tensile material testing, Figure 5-5 (Sanders et
al. 2004).
Liner Stiffness: These four liners were implemented as a 9 mm thickness layer.
Separate simulations were run with their increasing material stiffness (E ranging from
50050 to 88060 Pa) (Table 5-1). Simulations were modeled with both slip and no slip
conditions. All liners were modeled with a Poisson’s ratio of 0.49 which is based on silicon
and urethane polymer material properties and representative of the liner materials (O’Hara
1983; Noor & Mahmud 2015; Thomson 1966).

Figure 5-5. Tensile testing results for various prosthetic liners fell into four groups represented
by the liner boxed regions with decreasing stiffness (Daly & Odland 1979; Sanders et al. 2004).

Liner Thickness: After the gel liner stiffness changes were modeled, one of the
four liners, the Alpha liner was modeled at three gel thicknesses representing 3, 6 and 9
mm. These options are available by the manufacturer and different thicknesses are
commonly worn by amputees. The change of liner thickness was modeled to evaluate
the influence of liner thickness on underlying tissue stresses.
85

5.3 Results
5.3.1 Simplified Model Results
5.3.1.1 One and Two Layer Models
The purpose of the simplified models was to compare theory with model results. With a
distributed pressure of 10 Pa applied in the negative z direction, the model responded with a
compressive axial stress according to the linear elasticity theory (a negative sign indicated
compression). Theoretical calculations resulted in a -10 Pa compressive stress, calculated as 10
Pa distributed pressure divided by a one square meter cross section in the model. FE model
results yielded -9.946 Pa, matching well when evaluated as an average of elements in the center
of the model. When the distributed pressure was increased to 5 kPa, the model responded in the
nonlinear hyper-elastic manner as theoretically described in non-linear continuum mechanics for
neo-Hookean materials by Equation 2 (Bonet & Wood 2008).
𝜎=

𝐸
𝐽

ln(𝜆)

(2)

Where λ is stretch defined by the current length divided by the initial length; E is the young’s
modulus; J is the volume ratio defined as 𝜆(1−2𝑣) where 𝑣 is poisson’s ratio. 𝐽 equals one for
incompressible materials, however using our FE stretch and defined poisson’s ratio, 𝐽 was
calculated to be 0.999997. Therefore, also representing an incompressible material without
making the assumption that it was incompressible based on the volume ratio equation.
Theoretical calculations using Equation 2 resulted in -3.96 kPa for the 5 kPa load compared with
the -4.49 kPa stress in FEBio (Table 5-2). When providing numerical model data for all cases,
the stress was taken from the center of the analyzed layer as an average of four elements.

86

Table 5-2. Comparison results of two compressive pressures with FEA and theoretical axial
stress for a linear and non-linear response.
10 Pa Compressive Load
Linear Response
FEA
σz= -9.946 Pa
Theory σz= -10.0 Pa

5 kPa Compressive Load
Non-linear Response
FEA
σz = -4490 Pa
Theory σz = -3960 Pa

Figure 5-6. a) 10 Pa compressive load b) 5 kPa compressive load; both modeled as a single
layer neo-Hookean material model but note linear and non-linear response differences. (Scales
are in Pa.)

Next, simulations were run to compare the stress output of two material layers with a tied
interface. The two layer model in compression resulted in similar results to the single layer
model and theoretical values at 5 kPa distributed loading on the top surface (Table 5-3 and
Figure 5-7). Top and bottom layers were close in compressive stress values. However, the
bottom layer had slightly larger compressive stresses due to the fixed base which created a
stress concentration. The tied interface with a high penalty to prevent separation did not create
a noticeable stress concentration between the layers in the FEA output. Comparing top and
bottom layers, both had equal material properties and therefore stresses within both layers were
evenly distributed with similar results (Table 5-3).

87

Table 5-3. Compressive model results for a two layer neo-Hookean model with compressive
stress reported. Reported model outputs were evaluated in FEBio’s PostView software.
5 kPa Top layer
σz = -4490 Pa

5 kPa Bottom layer
σz = -4898 Pa

Figure 5-7. Compressive 5 kPa distributed pressure alone with tied interface between layers.
Bottom layer E = 50kPa; Top layer E = 50kPa (Equal thickness layers; White dashed line
represents contact interface between materials; Scales are in Pa)

Next a two layer model with uneven layer thicknesses was developed to investigate how
stresses could change when uneven layer thickness joined by a tied interface were introduced. A
second two layer model with uneven layer thickness and slightly different elastic moduli resulted
in similar stresses to previous models at 5 kPa distributed loading on the top surface (Table 5-4
and Figure 5-8).

Top and bottom layers were close in compressive stress values with 4.4 and

4.5 kPa on top and bottom respectively. Again, the tied interface between layers with a high
penalty to prevent separation also did not cause an influential difference at the contact for stress
analysis. The uneven thickness did not influence the results or skew model values compared
with expected non-linear continuum mechanics (Equation 2).

88

Table 5-4. Compressive model results for a two layer neo-Hookean model with compressive
stress reported. Comparison averaged elements from the middle of layers which resulted from
model outputs for the top and bottom layers.
5 kPa Top layer
σz = -4432 Pa

5 kPa Bottom layer
σz = -4506 Pa

Figure 5-8. Compressive 5 kPa distributed pressure alone with tied interface between layers.
Bottom layer E = 50kPa (0.8m); Top layer E = 40kPa (0.2m). (White dashed line represents
contact interface between materials; Scales are in Pa)

One and two layer models were developed for the purpose of understanding the effects
of geometry, interface contact definitions and uneven layer thicknesses.

These results

demonstrated minimal differences in stress surrounding a contact definition and good agreement
with theoretical calculations.

89

5.3.2 Realistic Four Layer Model Results
5.3.2.1 Cylindrical versus Rectangular
Modeling comparison results for an axis symmetric geometry versus the rectangular
model are reported here. Presented stresses are from an average of four centrally located
elements on the surface of each layer (Figure 5-9).

Figure 5-9. Four averaged elements were selected for the top surface of each layer. The same
placement of elements represented were consistent with deeper tissue layers.

Comparisons represented are for both combined loading (5 kPa pressure and 6 mm
displacement), as well as 5 kPa load alone (Table 5-5 and 5-6). The 6 mm displacement was
applied in the horizontal positive x direction, parallel to the top surface. Three stresses are
reported for 1) Von Mises, 2) Z normal compressive and 3) maximum shear. Von Mises stresses
are reported in FEBio as effective stresses but for engineering consistency will be labeled as Von
Mises in all results reported.

90

Table 5-5. Combined loading (5 kPa distributed pressure and 6 mm displacement) results for
cylindrical versus rectangular four layer model.
Material Layer
Gel Liner Cylinder
Gel Liner Square
Skin Cylinder
Skin Square
Muscle Cylinder
Muscle Square

Von Mises (kPa)
5.303
5.099
8.598
8.260
3.442
3.276

Z Normal Stress (kPa)
-4.626
-4.683
-5.328
-5.317
-4.987
-5.000

Max Shear Stress (kPa)
2.925
2.715
4.873
4.626
1.875
1.786

Figure 5-10. Axis Symmetric Comparison (10 cm x 10 cm Cube versus a cylinder with a radius
of 5.6419 cm; A = 0.01 m^2). Combined loading with 5kPa pressure and 6mm x direction
displacement. Tied interface with no slip condition and Willow wood Alpha Silicone Gel Liner.

91

Table 5-6. Results for cylindrical versus rectangular models with a 5 kPa distributed pressure
on top surface.
Material Layer
Gel Liner Cylinder
Gel Liner Square
Skin Cylinder
Skin Square
Muscle Cylinder
Muscle Square

Von Mises Stress (kPa)
5.121
4.753
8.692
8.492
1.826
1.835

Z Normal Stress (kPa)
-4.431
-4.484
-4.691
-4.658
-4.438
-4.448

Max Shear Stress (kPa)
2.546
2.370
4.347
4.250
0.914
0.917

Figure 5-11. Axis Symmetric Comparison (10 cm x 10 cm Cube versus a cylinder with a radius
of 5.6419 cm; A = 0.01 m^2). Loading with 5kPa pressure alone. Tied interface with no slip
condition and Willow wood Alpha Silicone Gel Liner. (Scales are in Pa)

92

5.3.2.2 Vertical Face Constraints

To evaluate the effects of the vertical constrains, results of two different four layer models
with tied interfaces between the skin and gel liner were compared. For comparison one model
had a zero displacement constraint on the xz vertical faces while on the other the vertical
constrains were removed (ie. front and rear surfaces) (Table 5-7). Overall without constraints,
the Von Mises stress increased, the normal stress increased at the bone to muscle interface, and
the shear stress increased.

Table 5-7. Model with and without vertical constraints on the xz face in positive and negative y
directions. Middle locations were defined as elements selected in the center of the layer’s cross
sectional view at the layer surface as seen in Figure 5-9.
Material Layer
Gel Liner w/vert
Gel Liner wo/vert
Skin w/vert
Skin wo/vert
Muscle w/vert
Muscle wo/vert

Von Mises Stress (kPa)
3.390
4.922
4.274
8.219
2.380
2.506

Z Normal Stress (kPa)
-5.105
-4.714
-5.942
-5.354
-5.480
-4.600

Max Shear Stress (kPa)
1.944
2.701
2.467
4.618
1.374
1.446

Figure 5-12. a) Model with vertical constraints on the xz face in positive and negative y
directions. b) Model with vertical constraints removed. (Scales are in Pa.)

93

5.3.2.3 Slip versus No Slip Model Results
In models where slip occurred between the gel liner and the surface of the skin, the Von
Mises stresses varied throughout the skin layer with the highest stresses occurring in the region
identified as element 2 (Figure 5-13a and 5-14); whereas models without slip (Figure 5-13a)
demonstrated even stresses throughout the interface contact. For the alpha liner, with identical
loading conditions, the Von Mises stress at the muscle to bone interface in the no slip model was
2.65 kPa compared to 0.39 kPa in the model with slip (Table 5-8). Additionally, maximum shear
stress was increased in all three locations for the model without slip with the highest value being
at location 2 equal to 5.09 kPa.

Figure 5-13: a) Gel layer hidden and presenting the model without slip, demonstrating two
locations of Von Mises stress analysis on the skin. b) Location 3 is in the middle y direction of
the muscle to bone interface. (Scales are in Pa.)

94

Figure 5-14. Gel layer hidden with difference is stress distribution in skin layer. a) Model with no
slip. b) Model with slip. (Scales are in Pa.)

Table 5-8: Comparison for slip versus no slip conditions in three locations (same liner used in
both). Location 1 and 2 are in the skin layer of the model and location 3 is at the muscle to bone
interface (Figure 5-14).
No Slip Model
Von Mises Stress (kPa)
Z Compressive Stress (kPa)
Max Shear Stress (kPa)

Slip Model

Location 1

Location 2

Location 3

Location 1

Location 2

Location 3

9.12
-4.78
4.98

9.24
-4.96
5.09

2.65
-4.63
1.53

6.71
-5.12
3.88

8.39
-4.80
4.84

0.39
-4.88
0.22

5.3.2.4 Liner Stiffness Results
The stresses caused by four liners were also compared. With increasing prosthetic gel
liner stiffness (i.e. Alpha to Iceross Comfort to DERMO to TEC Urethane), models without slip
demonstrated increased compressive stresses, and increased Von Mises and maximum shear
stresses at the bone to muscle interface. Compressive stresses in the skin were similar for all
liners. Maximum shear stresses were reduced in the skin as liner stiffness increased (Table 59). These were consistent trends for the models without slip. However, increasing gel liner
stiffness in models with slip were consistent with no increasing or decreasing trends.
95

Table 5-9. Comparison of Von Mises, normal compressive and maximum shear stress for four
different modeled prosthetic gel liners with model comparisons across a no slip and slip
condition of the liner with respect to the skin.
ALPHA Liner
(E = 50.05 kPa)
Von Mises Stress (kPa)
Z Compressive Stress (kPa)
Max Shear Stress (kPa)
Iceross Comfort
(E =55.86 kPa)
Von Mises Stress (kPa)
Z Compressive Stress (kPa)
Max Shear Stress (kPa)
DERMO
(E = 56.28 kPa)
Von Mises Stress (kPa)
Z Compressive Stress (kPa)
Max Shear Stress (kPa)
TEC Urethane
(E = 88.06 kPa)
Von Mises Stress (kPa)
Z Compressive Stress (kPa)
Max Shear Stress (kPa)

Location 1
9.120
-4.777
4.976

No Slip Model
Location 2 Location 3
9.236
2.653
-4.955
-4.625
5.090
1.531

No Slip Model
Location 1 Location 2 Location 3
8.848
8.813
2.722
-4.833
-5.040
-4.742
4.797
4.871
1.572

No Slip Model
Location 1 Location 2 Location 3
8.829
8.784
2.727
-4.837
-5.047
-4.75
4.785
4.855
1.574

No Slip Model
Location 1 Location 2 Location 3
7.846
7.172
2.908
-4.995
-5.385
-5.23
4.112
4.000
1.679

Location 1
6.687
-4.965
3.859

Slip Model
Location 2 Location 3
8.347
0.385
-4.924
-4.88
4.815
0.216

Location 1
6.687
-4.963
3.859

Slip Model
Location 2 Location 3
8.352
0.384
-4.927
-4.885
4.818
0.215

Location 1
6.687
-4.963
3.859

Slip Model
Location 2 Location 3
8.352
0.384
-4.927
-4.885
4.818
0.215

Location 1
6.664
-4.944
3.845

Slip Model
Location 2 Location 3
8.364
0.382
-4.933
-4.916
4.824
0.214

5.3.2.5 Liner Thickness Results
Stresses in underlying tissues were also evaluated for three gel liner thicknesses: 9 mm,
6 mm, and 3 mm (Table 5-10). For models with no slip, trends were seen. At location one, skin
Von Mises and maximum shear stresses increased with decreasing liner thickness. Location 2
was relatively unchanged for skin Von Mises and maximum shear stresses. For both locations 1
and 2, compressive stresses were unchanged. At muscle to bone interface (Location 3), Von
96

Mises and max shear stresses increased with decreasing liner thickness while compressive
stresses decreased (Table 5-10). For the models with slip, skin Von Mises stresses at location 1
were unchanged for varying gel liner thickness. For location 2, skin Von Mises, compressive
stresses and shear stresses decreased with thinner liners. At the bone to muscle interface,
compressive stresses decreased with thinner liners but Von Mises and shear stresses remained
unchanged.

Table 5-10. Comparison of TEC Urethane liner for three thicknesses of 9, 6 and 3 mm for no
slip and slip model definitions between the gel liner and skin interface.
TEC Urethane
(E = 88.06 kPa)

9 mm
Von Mises Stress (kPa)
Z Compressive Stress (kPa)
Max Shear Stress (kPa)
TEC Urethane
(E = 88.06 kPa)

6 mm
Von Mises Stress (kPa)
Z Compressive Stress (kPa)
Max Shear Stress (kPa)
TEC Urethane
(E = 88.06 kPa)

3 mm
Von Mises Stress (kPa)
Z Compressive Stress (kPa)
Max Shear Stress (kPa)

Location 1
7.846
-4.995
4.112

No Slip Model
Location 2 Location 3
7.172
2.908
-5.385
-5.230
4.000
1.679

Location 1
6.664
-4.944
3.845

Slip Model
Location 2 Location 3
8.364
0.382
-4.933
-4.916
4.824
0.214

No Slip Model
Location 1 Location 2 Location 3
7.999
7.053
3.112
-5.081
-5.339
-4.977
4.258
4.000
1.797

Slip Model
Location 1 Location 2 Location 3
6.650
8.349
0.376
-4.903
-4.881
-4.780
3.837
4.816
0.211

No Slip Model
Location 1 Location 2 Location 3
8.208
7.174
3.325
-5.063
-5.216
-4.72
4.479
4.117
1.920

Slip Model
Location 1 Location 2 Location 3
6.617
8.296
0.378
-4.863
-4.821
-4.735
3.819
4.785
0.212

5.4 Discussion
The presented work provides new analyses of skin and muscle tissue based on
experimental data inputs from prosthetic users. Further, the effects of liner stiffness and thickness
were evaluated as a result of various gel liner scenarios which are commonly worn by transtibial
amputees. In order to arrive at the complex four layer models, simplified models first investigated
97

the influences of various neo-hookean material responses to different geometries, constraints and
layers. The results can be used by engineers and clinicians to redesign material properties and
select appropriate liners given knowledge of the patient’s loading and displacement within the
socket.

5.4.1 Simplified Models
Simplified simulations were used to document the effects of geometry, layering and
constraints. Results of these models provided us with a basis for the more complex four layer
models. Stress results of simplified models corresponded well with non-linear continuum
mechanics theory. Next, one variable at a time was changed, such as the layers, material
properties and contacts.

These results supported the development of the four layer models by

understanding these interactions. Contacts between layers did not create a stress concentration
at the interface, guiding interpretation of stresses seen in gel, skin and muscle layers later.

5.4.2 Realistic Four Layer Models
5.4.2.1 Cylindrical versus Rectangular Geometry
In the four layer model, the layer thicknesses were selected based on the literature,
however, the region could have been represented as a circular or square cross-section because
the layer thicknesses were consistent throughout the individual layers. Therefore, a comparison
of the results for two geometries was conducted. The data between cylindrical and square
geometries were similar for for Von Mises and shear stresses, supporting the use of either
geometry for further models.

5.4.2.2 Vertical Face Constraints
For comparing a condition of slip between the liner and skin interface, a sliding with gaps
contact was implemented which required vertical face constraints. Therefore, on the models
98

which did not require vertical constraints, additional constraints were added and analyzed to
evaluate the effects of these constraints on stresses. The results demonstrated stresses with and
without vertical constraints in the tied interface contact definition. Tied interface models without
vertical constraints yielded tissue stresses higher than models with vertical constraints. For
example, Von Mises stresses at the bone to muscle interface decreased across constraint
conditions (no slip without constraints to no slip with constraints to slip with constraints). It can
be suggested that vertical constraints applied in the slip models do not account for the relative
change in tissue stresses and instead are associated with the contact interaction itself between
the skin and gel liner.

5.4.2.3 Slip versus No Slip
When comparing the gel interface contact models, a difference found was the increased
shear stress at the bone to muscle interface when not allowing slip to occur. Clinically prosthetic
liners were developed as a method to cushion the transfer of loads from the prosthetic socket to
the residual limb tissue (Klute et al. 2010). However, if the liner is slipping with respect to the skin
surface, the question arises of whether the interface is still providing the intended protection. Our
models demonstrate that compressive stresses are relatively similar across the slip versus no slip
models. Therefore, cushioning is being provided mechanically to the residual limb regardless of
slip. Yet, it is the increased shear at the bone to muscle interface that is of concern for deep
tissue ulcer formation (De Wert et al. 2015; Manorama et al. 2010; Manorama et al. 2013). As, it
has also been shown that with the addition of shear force, blood occlusion of underlying tissues
is increased when compared to a compressive load alone (Manorama et al. 2013). Further study
on the conditions of slip and no slip, particularly experimental studies are necessary.

99

5.4.2.4 Liner Stiffness
To further investigate the influences of gel liners on underlying tissue, four different gel
liners were modeled with increasing stiffness. Skin demonstrated a decreased maximum shear
stress with increasing stiffness for the models not allowing slip. Mechanically, the stiffest urethane
prosthetic liner is closest to mechanical properties of skin with a Young’s modulus of 88 kPa in
the liner versus 95 kPa for skin (Levy et al. 2015; Shoham et al. 2015). With maximum shear
stresses being the lowest for the urethane liner, it may suggest this would reduce skin irritation
when wearing this liner. The clinical intention is for the gel liner to grip the skin and move with
the residual limb. However, these stiffness results change when modeling the slip condition.
When slip occurred, it can be seen that maximum shear stress decreased in all simulations and
little change was observed between different liners. This is because once the liner slips relative
to the skin, it no longer transfers load differently based on the liner material stiffness. This would
suggest it is preferable to have a stiffer liner to minimize the level of shear stress in the skin.
Clinically, if a patient presents with skin irritation, a prosthetist often will change to a softer liner.
However, these model results suggest the opposite should be considered.

5.4.2.5 Liner Thickness
Prosthetic liner thickness is also commonly varied during prescription and development of
a custom device. Manufacturers often provide standard liner thicknesses such as 3, 6 and 9 mm.
Our models demonstrated little to no change in skin stresses but the bone to muscle interface
increased shear and Von Mises stress with decreasing liner thickness. Mechanically, with the
same amount of load and prescribed displacement it makes sense that a greater shear strain
would be experiences and resulting shear stress because the thicknesses of the overall model is
decreased.

100

5.4.3 Engineering Model Relevance to Influence Clinical Practices
With such a variety of prosthetic liner options, there is a developing need for engineering
based analyses to better understand and inform prosthetists of the mechanical influence of
various gel liners on the underlying tissue stresses. This work provides the needed mechanical
assessments of the gel liner interface.
Numerous studies have attempted to quantify skin material properties; however, it is
greatly dependent upon strain rate, load applied, type of tissue and location relative to the body
from which it was harvested (Gallagher, A. J.; Ní Annaidh, Aisling; Bruyère 2012; Sanders et al.
1995; Pailler-Mattei et al. 2008; Holt et al. 2008). Based on our experimental strain results, we
saw 6-11% strain. Walking is a cyclical event occurring approximately one cycle per second
(Perry, Jacquelin; Burnfield 2010). Skin ultimate tensile strength (UTS) of 0.25-1.0 MPa was
reported when loading at 0.25-10% strain per second at an initial stress of 10 kPa (Diridollou et
al. 2001; Zhou et al. 2010). These studies provide an appropriate comparison to our results
because we experimentally collected strain data within this range.

Based on the work of

numerous investigators, if loading is conducted at a slower strain rate of approximately 10% strain
per minute, the UTS increased to a range of 2-30 MPa (Silver et al. 2001; Ni et al. 2012; Gallagher
et al. 2012; Ankersen et al. 1999).
Simulations do not exceed published skin UTS values, however, the models represent a
single loading event instead of a fatigue evaluation to failure. Therefore, values need to be
considered relative to typical daily situations for prosthetic users.

On average, transtibial

amputees take 3395 steps per day (Stepien et al. 2007). Beyond a certain number of cyclical
loadings however, reparative mechanisms are exceeded and tissue breakdown begins. Future
studies could assess FEA fatigue simulations with repeated loading conditions.
Further evaluation of our models when comparing to literature and experimental data
supports the meaningful results of our reported stresses.
101

The skin layer experienced

approximately 9 kPa of Von Mises stress for the 9 mm Alpha liner, which compares well to typically
loaded skin samples which experienced 10 kPa of stress for in vitro testing (Diridollou et al. 2001).
Furthermore, gel liner shear stress of 2.7 kPa in the model, compares well with experimental
shear stress of 0.79-2.89 kPa during initial contact phase of walking found in our experimental
work.

5.4.4 Limitations, Future Work, Summary
A limitation of our models is that our FE analysis was conducted for one iteration of loading
and was not performed in a cyclical analysis. Consideration should be given to this for future
model developments so that the long term cyclical effects of load on the skin can be evaluated.
However, other limitations of the models presented include the generalized geometry of a block
region in the limb. The anatomical limb contains bony prominences and soft tissue thicker
regions. The changing bone geometry in the residual limb can create stress concentration
patterns because of the distinct differences in bone material properties. Future studies could
combine imaging techniques with our experimentally collected loading conditions to strengthen
model results dependent on anatomical differences.
Overall, these model results can provide insight into how soft-tissue responds due to
typical below knee amputee loading conditions with a modeled gel liner interface. Results suggest
that liner thickness and material properties should be considered when patients present with a
history of skin irritation or deep tissue injury based on documented trends in stresses throughout
these simulations. Commonly manufactured liners are typically one solid material. These results
could influence the further design of liners and consideration could include development of
functionally graded materials.

102

5.5 Conclusions

Soft-tissue finite element models are needed to improve our understanding of the interface
mechanics between prosthetic liners and underlying tissues in prosthetic users. Novel and unique
models were developed in a sequence of simulations to address questions regarding the gel liner
slip with respect to the skin and influences of liner stiffness and thickness on tissue stresses. The
comparison of slip and no slip conditions present an important clinical question because clinically,
slip of the liner with respect the limb is not desired. Additionally, if slip does not occur, the uneven
displacements on the liner would produce a shearing of deeper tissue which is also not desirable.
Trends reported for influences in tissue stresses for changing stiffness and thicknesses are an
important addition to the understanding of the interface mechanics. Amputees often present with
ulcers on their residual limb and added knowledge of stresses and strains can improve the
understanding of what is contributing to the wounds. Further model studies and experimental
data to support these models are warranted. Mechanics based data is a necessity to formulate
improved prosthetic fit and design, these additional data are also helpful in guiding prosthetists in
the selection of appropriate prosthetic liners for patients.

103

6.

CONCLUSIONS

104

6.1 Dissertation Conclusions
In conclusion, the completed research focused on quantifying mechanics occurring at the
socket to gel liner interfaced limb in transtibial amputees for better understanding of ulcer
formation. To experimentally assess the limb, a novel motion capture method was developed to
measure motion within the socket. The method was implemented in human subjects to evaluate
displacements and strains of the gel liner and with respect to the socket. Next a case study
measured shear and normal forces within the prosthetic socket wall during walking. Lastly,
previously collected experimental data was used for loading conditions in finite element models
to evaluate stresses of the underlying tissues due to typical walking loading conditions and various
gel liner interactions with the skin.
The first part of this research validated a methodology using motion capture to quantify
residual limb gel liner motion within the prosthetic socket.

Clear prosthetic sockets were

developed out of clinically used test socket material. Through the use of these sockets, motion
capture thin-disc markers were placed on meaningful locations of the residual limb on top of the
gel liner. These anatomical locations corresponded with soft tissue and bony landmark locations
which were intentionally positioned with respect to pressure tolerant and intolerant locations
where pressure ulcers commonly occur. Through numerous testing protocols, first static intermarker distances were captured with caliper and motion capture methods. For two residual limb
replicas, rigid and deformable, the analysis proved that inter-marker distances could be accurately
captured within the clear prosthetic socket. The test setup was also translated through space,
without changing the position of the limb with respect to the socket, to understand how different
angles of the limb relative to the motion capture cameras could calculate inter-marker distances.
Through statistical analyses, no differences were found for biplanar movement of the limb through
space, nor for the static measurements comparing caliper to motion capture measures. This
positively confirmed the ability for this method to be used to capture human residual limb and gel

105

liner motion during walking. Dynamic trials were completed for the replica limb mock-up which
demonstrated greater displacements distally near the pin on the gel liner insertion near the
suspension mechanism. This was further supported by a single case study for a transtibial
amputee during walking which analyzed greatest displacements distally in the residual limb. All
this data positively supported the use of this method in multiple transtibial amputees to evaluate
the influence of gel liner interface pin locking mechanisms on socket to limb interface movements.
The second part of this dissertation therefore implemented the previously validated
method in nine transtiblial limbs during walking. Four measures were calculated to evaluate the
various displacements and strains occurring within the prosthetic socket. Liner displacements
with respect to each other were calculated, displacements with respect to the socket, rotations
with respect to the socket and strains of the gel liner were all assessed during walking. Startling
data revealed large displacements at the distal end of the limb, likely due to the gel liner interface
with pin suspension insertion at the distal end of the limb. This data would indicate that users
with frequent limb irritation leading to wounds may be better suited with a suspension type that
does not cause increased strain at the distal end of the limb. Results also highlighted that limb
displacements were much greater than clinicians expected, therefore overestimating the
effectiveness of a “well fit prosthesis”. Results from this study should inform prosthetists of the
risks for transtibial amputees when experiencing pain, discomfort or wound formation over
commonly seen areas. Care should be given to redesigning sockets to minimize “play” within the
socket and therefore minimizing how much the limb can translate with respect to the socket during
walking.
The third experimental part included work measuring shear forces at the surface of the
skin and socket to limb interface. For three conditions in a single transtibial case study, shear
and normal forces were meaningfully analyzed during typical walking.

Patients often use

additional socks within their socket to account for volume changes and this study evaluated these
differences in addition to evaluating shear forces directly at the skin. Overall, this work was
106

important because it contributed to the understanding of shear force influence on changing socket
to limb forces. Many patients wear different interface thicknesses during the day and this case
study highlights a single sample of how these forces changed. The addition of shear force
measurement in addition to normal force is important for considering the underlying tissue
influences for blood perfusion and tissue health when additional shear is present. Finally, these
results are also helpful for creating finite element models which can be based on experimentally
captured data for boundary conditions and loading inputs.
The final part included finite element modeling of typical tissue thickness in the mid-fibular
region for a transitibal amputee. These models were constrained with previously discussed
experimental data to evaluate typical stresses occurring throughout the day in each step a limb
experiences. Additionally, two conditions were heavily analyzed for the conditions of whether the
gel liner slipped with respect to the skin surface or not as named, “slip and no slip conditions”.
Once these differences were understood, the models were modified to evaluate various gel liner
manufacturer properties such as stiffness and thickness to evaluate which liner selections would
be appropriate for particular cases based on the underlying tissue stress analyses.

With

increasing gel liner stiffness in the no slip models, skin maximum shear stress decreased and
bone to muscle compressive stress decreased. Thinner gel liners increased Von Mises and
maximum shear stresses in both skin and muscle tissues. Decreasing liner thickness also
decreased compressive stresses at the bone to muscle interface. Differences in stresses at the
superficial versus deep tissue can be helpful in understanding mechanisms leading to ulcer
formation. Soft-tissue finite element models are needed to improve our understanding of the
interface mechanics between prosthetic liners and underlying tissues in prosthetic users. The
comparison of slip and no slip conditions presented an important medical question because
clinically, slip of the liner with respect the limb is not desired. Additionally, if slip does not occur,
the uneven displacements on the liner would produce a shearing of deeper tissue which is also
not desirable. Further model studies and experimental data to support these models are
107

warranted. Mechanics based data is a necessity to formulate improved prosthetic fit and design,
these additional data are also helpful in guiding prosthetists in the selection of appropriate
prosthetic liners for patients.
In summary, this work provided 1) a new method used to quantify the most comprehensive
data set of within socket displacements, 2) interface shear and normal forces for a below knee
amputee which were used for modeling conditions, and 3) a series of models addressed tissue
stresses and liner interface mechanics. The presented research benefits the biomechanical world
by addressing multiple gaps in the literature which were evaluated using mechanical engineering
principles. This data can further understanding of pressure ulcer formation due to interface
mechanics and aid in understanding the fit of a prosthetic socket with a gel liner interface. The
combination of experimental data and modeling can be very powerful in the engineering world to
put emphasis on model importance when constrained with real world loading conditions.
Throughout this dissertation, the intention was to create new methods and analyses to improve
the life for amputees experiencing frequent pressure ulcers and wound formation. Since this area
is still misunderstood and an under analyzed topic, this work contributed to the literature to further
the improvement of socket fit, prosthetic device design and mechanics evaluation for transtibial
amputees.

108

APPENDICES

109

APPENDIX A: FE Modeling Review
Table A-1: Summary of FE modeling methodologies investigators implemented in the first
decade of prosthetic modeling. Original table, however, much of the content comes from a
published review article (Zachariah & Sanders 1996). E = Young’s modulus and v = Poisson’s
ratio.

110

APPENDIX B: FE Modeling with Imaging Review
Table B-1: Research summary FE models with improved imaging techniques but still using
linearly elastic material properties for biological tissues.

Investigator
(Year)

Source of
Geometric
data

Layers with
material
properties

Contacts
between layers

Boundary
Conditions

2D
or
3D

Verified with
Experimental
Data

Zhang,
Roberts
2000

Biplanar Xray views

3D

Yes, experimental
stress from load
cells

MRI

Contact,
separating and
slipping at skin to
liner
Unspecified

Weight bearing
in vertical
direction only

Lee, Zhang,
Jia, Cheung
2004
Lee, Zhang
2007

Soft tissues,
bone and liner:
isotropic and
linearly elastic
Linearly elastic,
isotropic,
homogeneous
Linearly elastic,
isotropic,
homogeneous

External surface
of socket fixed

3D

No

Soft tissue
boundaries
connect to bones
were fixed

3D

Yes, pressure of
painful regions

Zachariah,
Sanders
2000

CAD, CT,
MRI

Linear elastic,
isotropic,
homogeneous

Frictional slip
versus
frictionless slip at
socket

Bones fixed
boundaries
Force of load
tolerance
applied to each
site
800 N axial force
applied at distal
end of socket

3D

No

MRI

111

APPENDIX C: FE Modeling Using FEBio
Table C-1: Recent summary of FE models using FEBio in related biological research

Investigator
(Year)

Source of
Geometric
data

Layers with material
properties

Contacts
between layers

Boundary
Conditions

2D
or
3D

Verified with
Experimental
Data

Levy, Frank,
Gefen
2015

MRI

Bone and Achilies
tendon: linearelastic isotropic
Skin and fat: nearly
incompressible,
nonlinear isotropic
materials with large
deformation
behavior as
uncoupled neoHookean material

Tied interfaces
between all
tissue
components
Coefficient of
friction between
skin and support
0.43

Reaction force
at heel in WB
during supine
lying from
experimental
data

3D

No

Shoham,
Levy, Kopplin,
Gefen
2015

MRI

Tied interface
between all
tissues
Frictional sliding
between skin
and CFC

Body weight
along top
boundary.
Fixed surface
at inferior
edge of CFC

3D but
uniformly
extruded
from a
2D MRI
slice

No

Sengeh,
Moerman,
Petron, Herr
2016

MRI

Bone and contoured
foam cushion:
isotropic linear
elastic
Muscle, fat, skin:
nearly
incompressible and
hyperelastic
represented as a
neo-Hookean
material
Soft tissue (skinadipose layer and
internal muscle-soft
tissue complex):
non-linear elastic
(Mooney-Rivlin J = 1;
J is the determinant
of the deformation
gradient tensor) and
viscoelastic
materials

Not specified

Load curves
for loading
and unloading
applied at
each site
derived from
experimental
displacementtime data
Fixed at top
surface of
residuum

3D

Yes

Bones modeled as
rigidly supported
voids

112

APPENDIX D: IRB Approval Letter

113

APPENDIX E: IRB Consent Form

Measurement of Motions and Forces Associated with the Use of a
Lower Leg Prosthetic
Biomechanical Design Research Laboratory
and
Mary Free Bed Rehabilitation Hospital
Motion Analysis Center
You are being asked to participate in a biomechanics research project. The researchers are
required to provide a consent form to inform you about the study, to convey that participation is
voluntary, to explain risks and benefits of participation, and to empower you to make an
informed decision. You should feel free to ask the researchers any questions you may have.
The following provides an overview of testing that will occur in the research project entitled
“Measurement of Motions and Forces Associated with the Use of a Lower Leg Prosthetic”.
1. PURPOSE OF RESEARCH:
The purpose of this study is to measure the force patterns and tissue motion in the residual limb of
amputees. The results of this research work will help in understanding the development of pressure
ulcer sores in the prosthetic socket to limb interface.
2. WHAT YOU WILL BE ASKED TO DO:
If you consent to participate, the approximate duration of the testing is 3 hours and participation is on a
voluntary basis. An overview of the test will include, a short health questionnaire, sensors being applied
to your body to monitor your walking through the lab space, videotaping of your walking, photographs
of your leg, and a scan of your leg to create a 3D image. These tests will be performed twice on the first
day of participation and then repeated in a month.
First, you will be asked to complete a health questionnaire, then some basic physical measurements will
be taken such as height and weight and dimensions of the body segments and test regions. Certain
areas of the skin will be cleaned with an alcohol swab to assure that tape used for retro-reflective
markers will adhere. Hypo-allergenic medical tape will be used to attach these markers to the surface of
the skin. You may be asked to wear special clothing such as biking shorts or athletic tops so the sensors
can be positioned without the clothing interfering. We may ask that you wear clean laboratory clothing
if the attire you brought will not work with our sensors. Once all reflective markers are placed on the
body and your existing prosthetic device, you will be asked to perform walking trials through the
walkway of the laboratory. Motion and force data will be gathered and transferred electronically to a
computer. Secondly, a special instrumented prosthetic test socket will hold a force measurement device
while your body segment (residual limb) fits within the device as your typical prosthetic socket would fit
(customized to you). Once again you will walk multiple times though the designated walkway of the
laboratory. Motion and force data will again be gathered and transferred electronically to the
computer. Lastly, a clinically used tool will be used to scan your amputated limb in order to create a 3D
image of your leg. The scan is painless as no contact is required as a wand is merely waved around your
114

leg. The collected scan will be electronically transferred to the computer for anatomical and volume
measurement calculations.
This study contains multiple phases of data collection. The procedure will be repeated the same day in
the afternoon or evening (approximately 5 hours later). One month later the two repeated protocols
will be conducted in the morning and late afternoon/evening of the same day.
3. POTENTIAL BENEFITS:
You will not directly benefit from your participation in this study. However, your participation will help
in the gathering of data that will be useful in understanding the development of pressure ulcers and may
help identify better prevention strategies in amputees.
4. POTENTIAL RISKS:
There is minimal risk of injury from your participation in this research. All of the work related to this
project is non-invasive and will be conducted under the supervision of individuals who are experts with
the involved techniques. You may experience some redness on the skin due to the adhesive tape. This
redness usually disappears within a few hours, and lotion is available to you upon the removal of the
tape.
5. PRIVACY AND CONFIDENTIALITY:
Participation remains confidential, and confidentiality will be maintained with all data, which will be
protected to the maximum extent allowable by law. References to all data, including photographs will
be made by using a coded identification number that will not reveal your identity. Research records will
be kept in locked filing cabinets, on password encrypted computers and/or encrypted files in the
Mechanical Engineering Department at Michigan State University. Only the study researchers and HRPP
will have access to your identifiable information. Your data and photographs will be stored with the
investigator for three (3) years after completion of the project. Other members of the study team may
have access to your numerical data but they will not have access to your identifiable information.
Access of these data by students is for the purpose of data processing. Published experimental results
will not reveal your identity.
In addition to the measurements mentioned above, you may also be photographed by one of the listed
researchers or other qualified laboratory assistants. This information may be used for presentation
and/or publications relating to the research study.
6. YOUR RIGHTS TO PARTICIPATE, SAY NO, OR WITHDRAW:
Your participation in this study is voluntary. You may choose not to participate at all, or you may refuse
to participate in certain procedures. You are free not to answer certain questions, and you may
discontinue your participation at any time without penalty or loss of benefits. Whether you choose to
participate or not will have no effect on your medical care or treatment at Mary Free Bed Rehabilitation
Hospital. You will be told of any significant findings that develop during the course of the study that may
influence your willingness to continue to participate in the research.
Further, the researchers may elect to discontinue testing on any individual with or without cause at any
time, or to disqualify any individual for any reason at any time.

115

7. COSTS AND COMPENSATION FOR BEING IN THE STUDY:
You will not be compensated for your participation. There are no costs to participate. The testing time
is approximately 3 hours.
8. THE RIGHT TO GET HELP IF INJURED:
If you are injured as a result of your participation in this research project, Michigan State University will
assist you in obtaining emergency care, if necessary, for your research related injuries. If you have
insurance for medical care, your insurance carrier will be billed in the ordinary manner. As with any
medical insurance, any costs that are not covered or in excess of what are paid by your insurance,
including deductibles, will be your responsibility. The University's policy is not to provide financial
compensation for lost wages, disability, pain or discomfort, unless required by law to do so. This does
not mean that you are giving up any legal rights you may have. You may contact the Principal
Investigator, Tamara Reid Bush at (517) 353-9544 with any questions or to report an injury.
Further, your participation is voluntary, you may choose not to participate at all, or you may refuse to
participate in certain procedures or answer certain questions or discontinue your participation at any
time without penalty or loss of benefits. We appreciate your participation.
9. CONTACT INFORMATION FOR QUESTIONS AND CONCERNS
If you have concerns or questions about this study, such as scientific issues, how to do any part of it, or
to report an injury (i.e. physical, psychological, social, financial, or otherwise), please contact the
researcher:
Tamara Reid Bush, Ph.D.
Amy Lenz
2555 Engineering Building
East Lansing, MI 48824
Phone: 517-353-9544
Email: reidtama@msu.edu, lenzamy@msu.edu or amy.lenz@maryfreebed.com
If you have questions or concerns about your role and rights as a research participant, would like to
obtain information or offer input, or would like to register a complaint about this study, you may
contact, anonymously if you wish, the Michigan State University's Human Research Protection Program
at 517-355-2180, Fax 517-432-4503, or e-mail irb@msu.edu or regular mail at 408 W. Circle Drive, Room
207 Olds Hall, MSU, East Lansing, MI 48824.
10. DOCUMENTATION OF INFORMED CONSENT
Your signature below means that you voluntarily agree to participate in this research study.
I voluntarily agree to participate in this study.
 You are eligible to participate in the project even if you wish to not have photographs
taken.
o I agree to allow digital photos/ digital video clips during the experiment.
Yes

No

Initials____________

116

o

I agree to allow my photographs in reports and presentations without blocking
identifying features, i.e. it is ok to show your picture without a black box over your face
Yes
No
Initials____________

Signature:

Date:

This consent form was approved by a Michigan State University Institutional Review Board. Approved 01/05/17 – valid through
01/04/18. This version supersedes all previous versions. IRB # 14-089M.

117

BIBLIOGRAPHY

118

BIBLIOGRAPHY

Ali, S., Osman, N.A.A., et al., 2012. Clinical investigation of the interface pressure in the
trans-tibial socket with Dermo and Seal-In X5 liner during walking and their effect
on patient satisfaction. Clinical Biomechanics, 27(9), pp.943–948.
Ali, S. et al., 2013. Interface pressure in transtibial socket during ascent and descent on
stairs and its effect on patient satisfaction. Clinical Biomechanics, 28(9-10),
pp.994–999.
Ali, S., Abu Osman, N.A., et al., 2012. Qualitative study of prosthetic suspension
systems on transtibial amputees’ satisfaction and perceived problems with their
prosthetic devices. Archives of Physical Medicine and Rehabilitation, 93(11),
pp.1919–1923.
Ankersen, J. et al., 1999. Puncture resistance and tensile strength of skin simulants.
Journal of Engineering in Medicine, 213, pp.493–501.
Appoldt, F., Bennett, L. & Contini, R., 1969. Stump-socket pressure in lower extremity
prostheses. Journal of Biomechanics, 2(2), pp.207–208.
Arao, H., Shimada, T. & Hagisawa, S., 2013. Morphological characteristics of the
human skin over posterior aspect of heel in the context of pressure ulcer
development. Journal of Tissue Viability, 22(2), pp.42–51.
Armstrong, D.G. et al., 1997. Amputation and reamputation of the diabetic foot. Journal
of the American Podiatric Medical Association, 87(6), pp.255–259.
Attinger, C.E. et al., 2006. Angiosomes of the foot and ankle and clinical implications for
limb salvage: reconstruction, incisions, and revascularization. Plastic and
reconstructive surgery, 117(7 Suppl), p.261S–293S.
Bahr, C. et al., 1993. Lower-extremity amputation: incidence, risk factors, and mortality
in the Oklahoma Indian Diabetes Study. Diabetes, 42(6), pp.876–884.
Beil, T.L., Street, G.M. & Covey, S.J., 2002. Interface pressures during ambulation
using suction and vacuum-assisted prosthetic sockets. Journal of rehabilitation
research and development, 39(6), pp.693–700.
Bonet, J. & Wood, R.D., 2008. Nonlinear Continuum Mechanics for Finite Element
Analysis 2nd Editio., Cambridge, United Kingdom: Cambridge University Press.
Boulton, A.J.M., 2008. The diabetic foot: grand overview, epidemiology and
pathogenesis. Diabetes/metabolism research and reviews, 24 Suppl 1(December
2007), pp.S3–6.
Bouten, C. V. et al., 2003. The etiology of pressure ulcers: Skin deep or muscle bound?
Archives of Physical Medicine and Rehabilitation, 84(4), pp.616–619.
Boutwell, E. et al., 2012. Effect of prosthetic gel liner thickness on gait biomechanics
and pressure distribution within the transtibial socket. The Journal of Rehabilitation
119

Research and Development, 49(2), p.227.
Brechow, A. et al., 2013. Improving major amputation rates in the multicomplex diabetic
foot patient: focus on the severity of peripheral arterial disease. Therapeutic
advances in endocrinology and metabolism, 4(3), pp.83–94.
Brennan, J.M. & Childress, D.S., 1991. Finite element and experimental investigation of
above-knee amputee limb/prosthesis systems: a comparative study. ASME-BED,
20, pp.547–550.
Brown, B.J., Crone, C.G. & Attinger, C.E., 2012. Amputation in the diabetic to maximize
function. Seminars in vascular surgery, 25(2), pp.115–21.
Brunelli, S. et al., 2009. Vacuum assisted socket system in transtibial amputees: clinical
report. Orthopadie-Technik Quarterly, pp.2–7.
Burner, A. et al., 2013. Lifecasting in O&P. In In Proceedings of the 39th American
Academy of Orthotists and Prosthetists Annual Meeting. Orlando, FL.
Childers, W.L. & Siebert, S., 2015. Marker-based method to measure movement
between the residual limb and a transtibial prosthetic socket. Prosthetics and
orthotics international, pp.1–9.
Coleman, S. et al., 2013. Patient risk factors for pressure ulcer development:
Systematic review. International Journal of Nursing Studies, 50(7), pp.974–1003.
Collins, T.C. et al., 2002. Lower extremity nontraumatic amputation among veterans
with peripheral arterial disease: is race an independent factor? Medical care, 40(1
Suppl), pp.I106–I116.
Convery, P. & Murray, K.D., 2001. Ultrasound study of the motion of the residual femur
within a trans-femoral socket during daily living activities other than gait. Prosthetics
and orthotics international, 25(3), pp.220–7.
Convery, P. & Murray, K.D., 2000. Ultrasound study of the motion of the residual femur
within a trans-femoral socket during gait. Prosthetics and orthotics international,
24(3), pp.226–32.
Daly, C.H. & Odland, G.F., 1979. Age-related changes in the mechanical properties of
human skin. The Journal of Investigative Dermatology, 73(1), pp.84–87.
Dillingham, T.R., Pezzin, L.E. & Shore, A.D., 2005. Reamputation, mortality, and health
care costs among persons with dysvascular lower-limb amputations. Archives of
Physical Medicine and Rehabilitation, 86(3), pp.480–486.
Diridollou, S. et al., 2001. Skin ageing : changes of physical properties of human skin in
vivo. International Journal of Cosmetic Science, 23, pp.353–362.
Dombroski, S. et al., 2014. Silicone Model Project. In In Proceedings of the 40th
American Academy of Orthotists and Prosthetists Annual Meeting. Chicago, IL.
Dou, P. et al., 2006. Pressure distribution at the stump/socket interface in transtibial
amputees during walking on stairs, slope and non-flat road. Clinical Biomechanics,
120

21(10), pp.1067–1073.
Dudek, N.L., Marks, M.B. & Marshall, S.C., 2006. Skin problems in an amputee clinic.
American journal of physical medicine & rehabilitation / Association of Academic
Physiatrists, 85(5), pp.424–429.
Dumas, R., Cheze, L. & Frossard, L., 2009. Loading applied on prosthetic knee of
transfemoral amputee: Comparison of inverse dynamics and direct measurements.
Gait and Posture, 30(4), pp.560–562.
Edwards, M., 2000. Below knee prosthetic socket designs and suspension systems.
Physical medicine and rehabilitation clinics of North America, 11(3), pp.585–593.
Emrich, R. & Slater, K., 1998. Comparative analysis of below-knee prosthetic socket
liner materials. Journal of medical engineering & technology, 22(2), pp.94–98.
Erikson, U. & Lemperg, R., 1969. Roentgenological study of movements of the stump
within the prosthuis socket in below-knee amputees fitted. Acta orthop. Scandinav,
40, pp.520–529.
Eshraghi, A. et al., 2013. An experimental study of the interface pressure profile during
level walking of a new suspension system for lower limb amputees. Clinical
Biomechanics, 28(1), pp.55–60.
Fagundes, C. et al., 2005. Prognostic factors for amputation or death in patients
submitted to vascular surgery for acute limb ischemia. Vascular health and risk
management, 1(4), pp.345–9.
Fergason, J. & Smith, D.G., 1999. Socket considerations for the patient with a transtibial
amputation. Clinical Orthopaedics and Related Research, 361, pp.76–84.
Friberg, O., 1984. Biomechanical significance of the correct length of lower limb
prostheses: a clinical and radiological study. Prosthetics and orthotics international,
8, pp.124–129.
Gailey, R.S. et al., 2002. The Amputee Mobility Predictor: An instrument to assess
determinants of the lower-limb amputee’s ability to ambulate. Archives of Physical
Medicine and Rehabilitation, 83(5), pp.613–627.
Galbusera, F. et al., 2014. Comparison of various contact algorithms for poroelastic
tissues. Computer methods in biomechanics and biomedical engineering, 17(12),
pp.1323–34.
Galdik, J., 1955. The Below-Knee Suction Socket. Orthop Prosthet Appl J, 9, pp.43–46.
Gallagher, A.. J.., Ní Annaidh, A. & Bruyère, K., 2012. Dynamic Tensile Properties of
Human Skin. International Research Council on the Biomechanics of Injury, (IRC12-59), pp.494–502.
Gard, S.A., 2006. Use of quantitative gait analysis for the evaluation of prosthetic
walking performance. JPO: Journal of Prosthetics and Orthotics, 18(1), pp.93–104.
Gefen, A. et al., 2008. Strain-time cell-death threshold for skeletal muscle in a tissue121

engineered model system for deep tissue injury. Journal of Biomechanics, 41(9),
pp.2003–2012.
Gholizadeh, H. et al., 2015. A comparison of pressure distributions between two types
of sockets in a bulbous stump. Prosthetics and Orthotics International, pp.1–8.
Gholizadeh, H. et al., 2014. Clinical implication of interface pressure for a new
prosthetic suspension system. Biomedical engineering online, 13(1), p.89.
Gholizadeh, H. et al., 2014. Transtibial prosthesis suspension systems: Systematic
review of literature. Clinical Biomechanics, 29(1), pp.87–97.
Gholizadeh, H. et al., 2012. Transtibial prosthetic socket pistoning: Static evaluation of
Seal-In ÂŽ X5 and Dermo ÂŽ Liner using motion analysis system. Clinical
Biomechanics, 27(1), pp.34–39.
Golbranson, F.L. et al., 1988. Volume changes occurring in postoperative below-knee
residual limbs. Journal of Rehabilitation Research, 25(2), pp.11–18.
Haltmayer, M. et al., 2001. Impact of atherosclerotic risk factors on the anatomical
distribution of peripheral arterial disease. International Angiology, 20, pp.200–207.
Heim, M. et al., 1997. Silicone suspension of external prostheses. A new era in artificial
limb usage. The Journal of bone and joint surgery. British volume, 79, pp.638–640.
Henrot, P. et al., 2000. Imaging of the painful lower limb stump. Radiographics : a
review publication of the Radiological Society of North America, Inc, 20 Spec No,
pp.S219–S235.
Holman, K.H. et al., 2012. Racial Disparities in the Use of Revascularization Before Leg
Amputation in Medicare Patients. Journal of Vascular Surgery, 54(2), pp.420–426.
Holt, B., Tripathi, A. & Morgan, J., 2008. Viscoelastic response of human skin to low
magnitude physiologically relevant shear. Journal of Biomechanics, 41(12),
pp.2689–2695.
Johnson, K. & Davis, A.J., 2014. Lower Extremity Prosthetic Sockets and Suspension
Systems. In Prosthetic Restoration and Rehabilitation of the Upper and Lower
Extremity. New York, NY: Demos Medical Publishing, pp. 59–70.
Jude, E. et al., 2001. Peripheral Arterial Disease in Diabetic and Nondiabetic Patients.
Diabetes Care, 24(8), pp.1433–1437.
Kahle, J.T., 1999. Conventional and hydrostatic transtibial interface comparison. Journal
of Prosthetics and Orthotics, 11(4), pp.85–91.
Kapp, S., 1999. Suspension systems for prostheses. Clinical Orthopaedics and Related
Research, 361, pp.55–62.
Kapp, S.L. & Fergason, J.R., 2004. Transtibial Amputation: Prosthetic Management. In
Atlas of Amputations and Limb Deficiences: Surgical, Prosthetic and Rehabilitation
Principles. Rosemont, IL: American Academy of Orthopedic Surgeons, pp. 503–
515.
122

Khan, N.A. et al., 2014. Does the Clinical Examination Predict Lower Extremity
Peripheral Arterial Disease? Journal of the American Medical Association, 295(5),
pp.6–8.
Klute, G.K., Glaister, B.C. & Berge, J.S., 2010. Prosthetic liners for lower limb
amputees : A review of the literature. Prosthetics and Orthotics International,
34(June), pp.146–153.
Koc, E. et al., 2008. Skin problems in amputees: A descriptive study. International
Journal of Dermatology, 47(5), pp.463–466.
Krouskop, T. a et al., 1987. Computer-aided design of a prosthetic socket for an aboveknee amputee. Journal of rehabilitation research and development, 24(2), pp.31–
38.
Langewouters, G., Wesseling, K. & Goedhard, W., 1984. The static elastic properties of
45 human thoracic and 20 abdominal aortas in vitro and the parameters of a new
model. Journal of Biomechanics, 17(6), pp.425–435.
Le, K.M. et al., 1984. An In-depth look at pressure sores using monolithic silicone
pressure sensors. Plastic and reconstructive surgery, 74(6), pp.745 – 754.
Leavitt, L.A. et al., 1970. Quantitative method to measure the relationship between
prosthetic gait and the forces produced at the stump-socket interface. American
Journal of Physical Medicine and Rehabilitation, 49(3), pp.192–203.
Lee, W.C.C. et al., 2004. Finite element modeling of the contact interface between
trans-tibial residual limb and prosthetic socket. Medical Engineering and Physics,
26(8), pp.655–662.
Lee, W.C.C. & Zhang, M., 2007. Using computational simulation to aid in the prediction
of socket fit: A preliminary study. Medical Engineering and Physics, 29(8), pp.923–
929.
Levy, A., Frank, M.B.O. & Gefen, A., 2015. The biomechanical efficacy of dressings in
preventing heel ulcers. Journal of Tissue Viability, 24(1), pp.1–11.
Levy, S., 1995. Amputees: skin problems and prostheses. Cutis, 55(5), pp.297–301.
Levy, S. & Barnes, G., 1956. Verrucous hyperplasia of amputation stump. Archives of
Dermatology, 74(4), pp.448–449.
Liao, F., Burns, S. & Jan, Y., 2013. Skin blood flow dynamics and its role in pressure
ulcers. Journal of Tissue Viability, 22(2), pp.25–36.
Lilja, M., Johansson, T. & Oberg, T., 1993. Movement of the tibial end in a PTB
prosthesis socket: a sagittal X-ray study of the PTB prosthesis. Prosthetics and
orthotics international, 17(1), pp.21–6.
Lyon, C.C. et al., 2000. Skin disorders in amputees. Journal of the American Academy
of Dermatology, 42(3), pp.501–507.
Maas, S.A. et al., 2012. FEBio: Finite Elements for Biomechanics. Journal of
123

Biomechanical Engineering, 134(1), p.011005.
Mak, A.F.T., Zhang, M. & Tam, E.W.C., 2010. Biomechanics of Pressure Ulcer in Body
Tissues Interacting with External Forces during Locomotion. The Annual Review of
Biomedical Engineering, 12, pp.29–53.
Manorama, A. et al., 2013. Quantifying the effects of external shear loads on arterial
and venous blood flow: Implications for pressure ulcer development. Clinical
Biomechanics, 28(5), pp.574–578.
Manorama, A.A. et al., 2010. Blood perfusion and transcutaneous oxygen level
characterizations in human skin with changes in normal and shear loads Implications for pressure ulcer formation. Clinical Biomechanics, 25(8), pp.823–
828.
Marso, S.P. & Hiatt, W.R., 2006. Peripheral arterial disease in patients with diabetes.
Journal of the American College of Cardiology, 47(5), pp.921–929.
Meier, R.H., 1973. Stump-socket fit of below-knee prostheses: comparison of three
methods of measurement. Archives of Physical Medicine and Rehabilitation,
54(12), pp.553–558.
Meng, Q. et al., 2013. Comparison between FEBio and Abaqus for biphasic contact
problems. Proceedings of the Institution of Mechanical Engineers. Part H, Journal
of engineering in medicine, 227(9), pp.1009–19.
Miller, A., 2009. Gait event detection using a multilayer neural network. Gait and
Posture, 29, pp.542–545.
Muller, E.A. & Hettinger, T., 1952. A new method for measurement of pressure within
suction socket. Zeitschrift fur Orthopadie und Unfallchirurgie, 83(1), pp.132–137.
Naeff, M. & van Pijkeren, T., 1980. Dynamic pressure measurements at the interface
between residual limb and socket--the relationship between pressure distribution,
comfort, and brim shape. Bulletin of prosthetics research, 10(33), pp.35–50.
Narita, H., Yokogushi, K., Shii, S., Kakizawa, M. & Nosaka, T., 1997. Suspension effect
and dynamic evaluation of the total surface bearing (TSB) trans-tibial prosthesis: a
comparison with the patellar tendon bearing (PTB) trans-tibial prosthesis.
Prosthetics and orthotics international, 21(3), pp.175–178.
Neumann, E.S. et al., 2013. Use of a load cell and force-moment curves to compare
transverse plane moment loads on transtibial residual limbs: A preliminary
investigation. Prosthetics and orthotics international, 38(3), pp.253–262.
Newton, R.L. et al., 1988. Skeletal Radiology Radiological evaluation of prosthetic fit in
below-the-knee amputees. Skeletal Radiology, 17, pp.276–280.
Ni, A.A. et al., 2012. Characterization of the anisotropic mechanical properties of
excised human skin. Journal of Bechanical Behaviour of Biomedical Materials, 5,
pp.139–148.

124

Nixon, J. et al., 2006. Pressure relieving support surfaces: a randomised evaluation.
Health technology assessment (Winchester, England), 10(22), pp.2–6.
Noor, S.A. & Mahmud, J., 2015. Modelling and Computation of Silicone Rubber
Deformation Adapting Neo-Hookean Constitutive Equation. International
Conference on Communication Systems and Network Technologies, pp.1323–
1326.
O’Hara, G.P., 1983. Mechanical Properties of Silicone Rubber in a Closed Volume.
Technical Report ARLCB, 2(December).
Oomens, C.W.J. et al., 2016. How does lateral tilting affect the internal strains in the
sacral region of bed ridden patients? — A contribution to pressure ulcer prevention.
Clinical Biomechanics, 35, pp.7–13.
Owings, M.F. & Kozak, L.J., 1998. Ambulatory and inpatient procedures in the United
States, 1996. Vital and health statistics. Series 13, Data from the National Health
Survey, (139), pp.1–119.
Pailler-Mattei, C., Bec, S. & Zahouani, H., 2008. In vivo measurements of the elastic
mechanical properties of human skin by indentation tests. Medical Engineering and
Physics, 30(5), pp.599–606.
Papaioannou, G. et al., 2010. Assessment of amputee socket-stump-residual bone
kinematics during strenuous activities using Dynamic Roentgen Stereogrammetric
Analysis. Journal of biomechanics, 43(5), pp.871–8.
Pearson, J.R., 1974. Pressure variation in the below-knee, patellar tendon bearing
suction socket prosthesis. Journal of Biomechanics, 7(6), pp.487–496.
Perduta, A. & Putanowicz, R., 2015. Comparison of elastomer modelling in various FEM
packages. Procedia Engineering, 108, pp.510–517.
Perry, Jacquelin; Burnfield, J.M., 2010. Gait Analysis Normal and Pathological Function
2nd Editio., Thorofare, NJ: SLACK Incorporated.
van Pijkeren, T., Naeff, M. & Kwee, H.H., 1980. A new method for the measurement of
normal pressure between amputation residual limb and socket. Bulletin of
prosthetics research, 10(33), pp.31–34.
Pirouzi, G. et al., 2014. Review of the socket design and interface pressure
measurement for transtibial prosthesis. Scientific World Journal, 2014.
Portnoy, S. et al., 2011. Effects of sitting postures on risks for deep tissue injury in the
residuum of a transtibial prosthetic-user: a biomechanical case study. Computer
methods in biomechanics and biomedical engineering, 14(11), pp.1009–19.
Portnoy, S. et al., 2008. Internal mechanical conditions in the soft tissues of a residual
limb of a trans-tibial amputee. Journal of Biomechanics, 41(9), pp.1897–1909.
Portnoy, S., Siev-Ner, I., Shabshin, N., et al., 2009. Patient-specific analyses of deep
tissue loads post transtibial amputation in residual limbs of multiple prosthetic
125

users. Journal of Biomechanics, 42(16), pp.2686–2693.
Portnoy, S. et al., 2007. Real-time patient-specific finite element analysis of internal
stresses in the soft tissues of a residual limb: A new tool for prosthetic fitting.
Annals of Biomedical Engineering, 35(1), pp.120–135.
Portnoy, S. et al., 2010. Real-time subject-specific analyses of dynamic internal tissue
loads in the residual limb of transtibial amputees. Medical Engineering and Physics,
32(4), pp.312–323.
Portnoy, S., Siev-Ner, I., Yizhar, Z., et al., 2009. Surgical and morphological factors that
affect internal mechanical loads in soft tissues of the transtibial residuum. Annals of
Biomedical Engineering, 37(12), pp.2583–2605.
Pritham, C.H., 1979. Suspension of the below knee prosthesis: An overview. Ortho
Pros, 33(22), pp.1–19.
Prompers, L. et al., 2007. High prevalence of ischaemia, infection and serious
comorbidity in patients with diabetic foot disease in Europe. Baseline results from
the Eurodiale study. Diabetologia, 50(1), pp.18–25.
Prosthetists, A.A. of O. and, 2004. Post-Operative Management of the Lower Extremity
Amputee: Official Findings of the Consensus Conference. In American Academy of
Orthotists and Prosthetists.
Quesada, P. & Skinner, H., 1991. Analysis of a below-knee patellar tendon-bearing
prosthesis : A finite element study. Journal of Rehabilitation Reasearch, 28(3),
pp.1–12.
Radcliffe, C.W. & Foort, J., 1961. Patellar-tendon-bearing Below-knee Prosthesis 1st
ed., Berkeley, CA: University of California.
Reynolds, D.P. & Lord, M., 1992. Interface load analysis for computer-aided design of
below-knee prosthetic sockets. Medical and Biological Engineering and Computing,
30(4), pp.419–426.
Robbins, J.M. et al., 2008. Mortality rates and diabetic foot ulcers: is it time to
communicate mortality risk to patients with diabetic foot ulceration? Journal of the
American Podiatric Medical Association, 98(6), pp.489–493.
Sanders, J.E. et al., 2006. A noncontact sensor for measurement of distal residual-limb
position during walking. Journal of rehabilitation research and development, 43(4),
pp.509–516.
Sanders, J.E. et al., 2005. Changes in interface pressures and shear stresses over time
on trans-tibial amputee subjects ambulating with prosthetic limbs: Comparison of
diurnal and six-month differences. Journal of Biomechanics, 38(8), pp.1566–1573.
Sanders, J.E. et al., 2009. Clinical utility of in-socket residual limb volume change
measurement: case study results. Prosthetics and orthotics international, 33(4),
pp.378–390.

126

Sanders, J.E. et al., 1998. Effects of alignment changes on stance phase pressures and
shear stresses on transtibial amputees: Measurements from 13 transducer sites.
IEEE Transactions on Rehabilitation Engineering, 6(1), pp.21–31.
Sanders, J.E. et al., 2000. Effects of changes in cadence, prosthetic componentry, and
time on interface pressures and shear stresses of three trans-tibial amputees.
Clinical Biomechanics, 15(9), pp.684–694.
Sanders, J.E. et al., 1997. Interface pressures and shear stresses at thirteen socket
sites on two persons with transtibial amputation. Journal of rehabilitation research
and development, 34(1), pp.19–43.
Sanders, J.E. et al., 1998. Material properties of commonly-used interface materials and
their static coefficients of friction with skin and socks. Journal of rehabilitation
research and development, 35(2), pp.161–176.
Sanders, J.E. et al., 2004. Testing of elastomeric liners used in limb prosthetics:
classification of 15 products by mechanical performance. Journal of rehabilitation
research and development, 41(2), pp.175–186.
Sanders, J.E. & Daly, C.H., 1993. Normal and shear stresses on a residual limb in a
prosthetic socket during ambulation: Comparison of finite element results with
experimental measurements. Journal of Rehabilitation Reasearch, 30(2), pp.191–
204.
Sanders, J.E., Daly, C.H. & Burgess, E.M., 1992. Interface shear stresses during
ambulation with a below-knee prosthetic limb. Journal of rehabilitation research and
development, 29(4), pp.1–8.
Sanders, J.E. & Fatone, S., 2011. Residual Limb Volume Change: Systematic Review
of Measurement and Management. Journal of Rehabilitation Reasearch device,
48(8), pp.949–986.
Sanders, J.E., Goldstein, B.S. & Leotta, D.F., 1995. Skin response to mechanical
stress: Adaptation rather than breakdown - A review of the literature. Journal of
Rehabilitation Research and Development, 32(3), pp.214–226.
Schiff, a. et al., 2014. Quantification of Shear Stresses Within a Transtibial Prosthetic
Socket. Foot & Ankle International, 35(8), pp.779–782.
Schwartz, M.H. & Rozumalski, A., 2008. The gait deviation index: A new comprehensive
index of gait pathology. Gait and Posture, 28(3), pp.351–357.
Schwarze, M. et al., 2013. Loads on the prosthesis-socket interface of above-knee
amputees during normal gait: Validation of a multi-body simulation. Journal of
Biomechanics, 46(6), pp.1201–1206.
Sengeh, D.M. et al., 2016. Multi-material 3-D viscoelastic model of a transtibial
residuum from in-vivo indentation and MRI data. Journal of the Mechanical
Behavior of Biomedical Materials, 59, pp.379–392.
Shoham, N. et al., 2015. Contoured Foam Cushions Cannot Provide Long-term
127

Protection Against Pressure-Ulcers for Individuals with a Spinal Cord Injury:
Modeling Studies. Advances in skin & wound care, 28(7), pp.303–316.
Silver, F.H., Freeman, J.W. & Devore, D., 2001. Viscoelastic properties of human skin
and processed dermis. Skin Research and Technology, 7(9), pp.18–23.
Silver-Thorn, M.B., 1991. Estimation and experimental verification of residual
limb/prosthetic socket interface pressures for below-knee amputees. Northwestern
University.
Spires, M.C., Kelly, B.M. & Davis, A.J., 2014. Prosthetic Restoration and Rehabilitation
of the Upper and Lower Extremity 1st ed., New York, NY: demosMedical.
Steege, J.W. & Childress, D.S., 1988. Finite element estimation of pressure at the
below-knee socket interface by finite element analysis, Seattle, WA.
Stepien, J.M. et al., 2007. Activity Levels Among Lower-Limb Amputees : Self-Report
Versus Step Activity Monitor. Archives of Physical Medicine and Rehabilitation,
88(July), pp.896–900.
Taylor, G.I. & Pan, W.R., 1998. Angiosomes of the Leg: Anatomic Study and Clinical
Implications. Plastic and reconstructive surgery, 102(3), pp.599–616.
Tepole, A.B. et al., 2011. Growing skin: A computational model for skin expansion in
reconstructive surgery. J Mech Phys Solids, 59(10), pp.2177–2190.
Thomson, K.C., 1966. On the Complex Poisson ’ s Ratio of a Urethane Rubber
Compound. Journal of Applied Polymer Science, 10, pp.1133–1136.
De Wert, L.A. et al., 2015. A new method to evaluate the effects of shear on the skin.
Wound Repair and Regeneration, 23(6), pp.885–890.
Wolf, S.I. et al., 2009. Pressure characteristics at the stump/socket interface in
transtibial amputees using an adaptive prosthetic foot. Clinical Biomechanics,
24(10), pp.860–865.
Zachariah, S.G. et al., 2004. Shape and volume change in the transtibial residuum over
the short term: Preliminary investigation of six subjects. journal of Rehabilitation
Research and Development, 41(5), pp.683–694.
Zachariah, S.G. & Sanders, J.E., 2000. Finite element estimates of interface stress in
the trans-tibial prosthesis using gap elements are di ! erent from those using
automated contact. , 33.
Zachariah, S.G. & Sanders, J.E., 1996. Interface mechanics in lower-limb external
prosthetics: A review of finite element models. IEEE Transactions on Rehabilitation
Engineering, 4(4), pp.288–302.
Zhang, M. et al., 1995. Development of a non-linear finite element modelling of the
below-knee prosthetic socket interface. Medical Engineering and Physics, 17(8),
pp.559–566.
Zhang, M., Mak, A.F.T. & Roberts, V.C., 1998. Finite element modelling of a residual
128

lower-limb in a prosthetic socket: A survey of the development in the first decade.
Medical Engineering and Physics, 20(5), pp.360–373.
Zhang, M. & Roberts, C., 2000. Comparison of computational analysis with clinical
measurement of stresses on below-knee residual limb in a prosthetic socket.
Medical Engineering and Physics, 22(9), pp.607–612.
Zhou, B.B. et al., 2010. Strain rate sensitivity of skin tissue under thermomechanical
loading. Philosophical Transactions of the Royal Society, 368, pp.679–690.
Ziegler-Graham, K. et al., 2008. Estimating the Prevalence of Limb Loss in the United
States: 2005 to 2050. Archives of Physical Medicine and Rehabilitation, 89(3),
pp.422–429.
Zollner, A.M. et al., 2013. Growth on demand: Reviewing the mechanobiology of
stretched skin. J Mech Phys Solids, 28, pp.495–509.

129