Journal of Rehabilitation Research and Development

Vol. 39 No. 1, January/February 2002

Pages 1–11

Transtibial energy-storage-and-return prosthetic devices: A

review of energy concepts and a proposed nomenclature

Brian J. Hafner, BS; Joan E. Sanders, PhD; Joseph M. Czerniecki, MD; John Fergason, CPO
Department of Bioengineering, Department of Rehabilitation Medicine, University of Washington, Seattle, WA;
Seattle VA Medical Center, 1660 S. Columbian Way, Seattle, WA

Abstract—Prosthetic devices that can store and return energy INTRODUCTION

during gait enhance the mobility and functionality of lower-
limb amputees (1–4). The process of selecting and fitting such
Prosthetic feet that can store and release energy dur-
devices is complicated, partly because of confusing literature
ing gait can be beneficial to lower-limb amputees (1–4).
on the topic. Gait analysis methods for measuring energy char-
Current foot designs provide a wide range of perfor-
acteristics are often incomplete, leading to inconsistencies in
mance choices and, when fit appropriately, can improve
the energy classifications of different products. These inconsis-
tencies are part of the reason for the lack of universally accu-
the comfort and performance of a prosthetic limb. How-
rate terminology in the field. Inaccurate terminology ever, current designs also place additional demand on the
perpetuates misunderstanding. In this paper, important pros- rehabilitation team. The team must be familiar with the
thetic energy concepts and methods for measuring energy char- spectrum of available components, as well as how each
acteristics are reviewed. Then a technically accurate nomen- component might alter patient function. There are thus
clature and a method of functional classification are proposed. two requirements of clinicians for effective prosthetic
This review and proposed classification scheme should help to selection and for fitting of advanced prosthetic feet:
alleviate confusion and should facilitate enhancement of the understanding the principles of energy transfer, and
design, selection, and fitting of prosthetic limbs for amputee understanding how these devices differ.
patients. Unfortunately, the literature related to energy trans-
fer and prosthetic componentry is confusing. One prob-
lem is the variation in the methods used to measure the
Key words: gait analysis, lower-limb amputee, prosthetic feet. energy-storage and the energy-return features. Most
methods measure only part of the total energy character-
istics. A second problem, one that stems from the first, is
the lack of a universally common and technically accu-
This project is based upon work supported in part by the Whi- rate terminology. The result is a confusing literature that
taker Foundation, the National Institute of Child & Human
Development (grant number HD-31445), and the Department of confounds component selection and fitting.
Veterans Affairs.
The purpose of this review is to:
Address all correspondence and requests for reprints to Joan E. Sanders,
PhD, Department of Bioengineering, 357962, Harris Hydraulics 309, 1. Clearly explain the energy concepts and terms rel-
Seattle, WA 98195; jsanders@u.washington.edu. evant to energy transfer in prosthetics,
1
2

Journal of Rehabilitation Research and Development Vol. 39 No. 1 2002

2. Review methods for measurement of prosthesis

energy storage and energy return and discuss
which parts of the energy capabilities they mea-
sure, and
3. Propose a technically accurate nomenclature and
method of functional classification for prosthetic
devices that can store and release energy.

Figure 1.
BACKGROUND Crosssections of various energy-storing feet. Each foot is composed
of a compressible heel and flexible keel spring. A. Seattle Foot, B.
Dynamic Foot, C. STEN Foot, D. SAFE Foot, E. Carbon Copy II
One of the most important goals of rehabilitation fol- Foot. Image based on the images of Wing DC, Hittenberg DA. Arch
lowing a transtibial amputation is to return an individual Phys Med Rehabil 1989;70(4):330–5.
to the highest functional level of ambulation possible. A
successful rehabilitation involves a comprehensive pro-
cess of obtaining an optimum socket design, alignment, and clinical applications for each of the above feet have
and choice of prosthetic componentry. Prior to the early received prior attention and publication (5–7).
1980s, most prosthetic feet were designed with the goal In 1987, a radically different type of prosthetic
of restoring basic walking and simple occupational tasks. device was introduced into the market. The Flex-Foot
Active or athletic amputees, however, demand more than (Flex-Foot, Inc., Aliso Viejo, CA; Figure 2A) prosthesis
this minimum “functional level” of ambulation from their includes both a flexible carbon fiber shank and a heel
prostheses. These individuals have the additional goals of spring, which allow the entire length of the prosthesis,
being able to run, jump, and participate in sports. The rather than solely the foot, to flex, absorb, and return
demand for prostheses capable of higher levels of perfor- energy to the amputee. This unconventional design is
mance shaped the design and manufacture of the so- considered by many to be the most “advanced” energy-
called “energy storing” foot, a foot capable of storing storing prosthetic device available. Newer and more
energy during stance and returning it to the amputee to sophisticated prosthetic designs such as the Reflex VSP®
assist in forward propulsion in late stance. This foot (Flex-Foot, Inc.; Figure 2B) may continue to improve on
design was met with great clinical success and soon the performance of the Flex-Foot, but have received little
became a driving force in the design of prosthetic feet. attention in the literature thus far (8,9). In 1988, another
The introduction of the Seattle Foot™ in 1981 design similar to the Flex-Foot, was developed by
brought about the inception of the first so-called “energy- Springlite (Salt Lake City, UT). The Springlite Advan-
storing” prosthetic foot (ESPF). The Seattle Foot (Seattle tage DP foot (Figure 2C) utilizes a carbon/epoxy pylon
Limb Systems, Poulsbo, WA) incorporates a flexible that flexes under the weight of the amputee but is a
Delrin® (DuPont, Wilmington, DE) keel inside a poly- unique one-piece design (the heel spring is fused to the
urethane shell. It is this Delrin keel that flexes during pylon spring with a compressible urethane elastomer heel
loading, acting as an elastic spring, returning a portion of web). The Springlite foot, while a commonly used clini-
the input energy to the amputee later in gait. cal device, has received little attention in the literature. In
Other feet followed a pattern similar to the Seattle 2000, another energy-storing foot was introduced. The
Foot and incorporated a flexible keel(s) surrounded by Ohio Willow Wood Pathfinder® is similar in concept to
foam and/or a polyurethane cosmesis. Such feet include the Flex-Foot Reflex VSP but adds an adjustable heel
the Dynamic (Otto Bock Industries, Minneapolis, MN), shock absorber to a composite keel spring system (Fig-
STEN (Kingsley Manufacturing Co., Costa Mesa, CA), ure 2D). Such a design allows the foot to be specifically
SAFE (Campbell-Childs, Inc., White City, OR), Carbon “tailored” to the activity level and task of the amputee.
Copy II (Ohio Willow Wood Co., Mount Sterling, OH), As prosthetic devices become more complex, the
TruStep® (College Park Industries, Inc., Fraser, MI), need for understanding the mechanical performance of
Quantum (Hanger Orthopedic Group, Bethesda, MD), prostheses becomes ever more critical. Ultimately, both
and others (Figure 1). Specific construction differences ESPF and conventional prosthetic feet are passive devices
 3

HAFNER et al. Transtibial prosthetic devices: Energy concepts

Figure 2.
Advanced energy-storing prostheses: A. Modular III, B. Reflex VSP,
C. Advantage DP, and D. Pathfinder. Figure 3.
Potential energy derived from a spring in compression corresponds to
area A.
and, as such, will never fully attain the performance of the
unamputated limb (an active system with muscular forces
Elasticity vs. Viscoelasticity
and sensory feedback). Despite this limitation, there have
A compressed (theoretical) elastic spring will return
been significant advances in the devices themselves that
100 percent of the potential energy as work when it is
may greatly improve the performance and the activity released. This theoretical energy is called the elastic
level of the amputee. To better evaluate and analyze the potential energy of the spring. An elastic spring will
performance of such devices, one must understand the return to its original shape via the same path that was
basic principles upon which they have been designed and used to compress it, as shown in Figure 3. In reality, no
engineered. spring is 100 percent efficient. Rather than return to its
original state via the same path on the force-deformation
curve as when it was compressed, a real spring will
Energy Concepts
return via a different path because of friction in the spring
and energy lost as heat and/or sound. This behavior,
Principles of Energy Storage called viscoelasticity, is identified by hysteresis, the dif-
The relationship between work and energy is a fairly ference between the loading and unloading portions of
simple one, yet the two terms are many times used inter- the load-deformation curve (Figure 4).
changeably in the literature surrounding ESPF. Energy is The energy lost in this system as a result of friction
the capability of a material to do work. In the ideal case, is equivalent to area B between the two curves (i.e., the
the energy and the work of an object are identical, but in area under the loading curve minus the area under the
unloading curve) and is dissipated as heat and sound.
reality, the work an object performs is always less than
This area between the input and output curves is also
the stored energy it possesses because of heat, sound, and
known as the dissipated energy and is equivalent to the
other losses. For simplicity, consider the prosthesis as a input energy minus the output energy:
simple mechanical spring (in reality, it is more accurately
described as a system of springs and other mechanical
components). During gait, work is provided by the
weight of the body to load the spring into compression.
Energy, as denoted by many prosthetics researchers
The material of the prosthesis (i.e., the spring) then stores
when describing an ESPF, is simply the work input to the
this work as potential energy and can release it as work to
prosthesis during different phases of gait. The energy
act upon another object when the compressive force is stored and returned by a prosthesis is typically calculated
released. Work is calculated by integration of the force- by integrating under the ankle power-time curve measured
deformation curve generated by compression of the with gait analysis equipment, a quantity that approximates
spring (Figure 3). The potential energy of the com- the energy measurement derived from a force-deforma-
pressed spring corresponds to area A under the curve. tion curve. These joint powers are calculated across each
4

Journal of Rehabilitation Research and Development Vol. 39 No. 1 2002

sor hallicus longus, and extensor digitorum longus)

eccentrically contract to absorb shock and provide con-
trolled plantarflexion of the foot. This action continues
until the foot is flat on the ground at the end of the load-
ing response. The plantarflexors (soleus, gastrocnemius,
flexor digitorum longus, flexor hallucis longus, and pero-
neus longus and brevis) then eccentrically contract dur-
ing controlled rotation of the tibia over the foot before
concentrically contracting to propel the limb forward and
initiate heel-rise. This final concentric action provides the
primary power in the ankle during gait.
In quantitative gait analysis evaluations, the ankle
has been shown to produce substantially more work than
any other joint in the lower limb (10,12). In a study of
nine normal subjects at self-selected walking velocity,
the ankle joint muscles produced an average of 540 per-
cent more work than they absorbed during gait (10). This
active generation of power is critical to the production of
Figure 4.
Simulated load-deformation curve of a viscoelastic material. B = natural gait. Effective replacement of this power genera-
energy lost as a result of friction. tion is one of the major barriers to total gait replication
with a prosthetic system.

joint by motion analysis software with force plate and

kinematic data because the total deformation of the spring Energy in Lower-Limb Prosthesis
cannot be measured directly. The goal of complete physiological replacement of
an amputated foot and ankle with a prosthetic device is
an ambitious and, as of yet, unattained aim. Because the
Energy in Intact Ankle Joint
musculoskeletal complex of the foot and ankle not only
The ankle complex provides most of the work pro-
absorbs energy but also generates more energy than it
duced during gait (10). The ankle joint is formed by a
absorbs, active prosthetic components would be required
highly sophisticated system of bones, muscles, tendons,
to completely replace the lower limb. However, current
and ligaments. The predominant motion of the ankle joint
in walking gait is in the sagittal plane, and the majority of commercial prostheses are composed of passive materi-
gait analysis techniques developed (and those discussed als, and thus can at best only partially replace the missing
in this paper) are focused on that plane of motion. For physiological system. This leads to marked asymmetries
simplicity, the ankle joint is often analyzed with the use in the temporal (13–20), kinetic (2,17,20–24), and kine-
of the link-segment model that represents the leg bones matic (14,19,21,23–25) gait parameters between each
and the foot as two rigid segments on either side of an limb of an amputee during walking gait.
articulating joint (11). The two primary muscle groups of The prosthetic foot-ankle complex achieves its par-
the ankle, the plantarflexors and dorsiflexors, govern the tial replacement of the energy features of the normal
relative motion between these rigid bodies. This model, physiological system through two main components, the
while technically inaccurate (the foot is actually com- heel and the keel. Both components can absorb shock and
posed of 26 individual bones), sufficiently represents the store and release energy. However, in general, the heel
gross motion of the ankle joint for most analysts’ functions primarily as an energy-absorbing mechanism
purposes. when the limb strikes the ground at initial contact. The
In the function of the intact ankle, the muscles of the keel functions as a stable surface for stance and, in some
leg provide the majority of shock absorption, controlled prostheses, such as the Seattle or Flex-Foot, as a propel-
motion, and power generation (Figure 5). Upon ground ling mechanism to push the amputee into the next step of
contact, the primary dorsiflexors (tibialis anterior, exten- gait. Together, the heel and keel both absorb and return
 5

HAFNER et al. Transtibial prosthetic devices: Energy concepts

energy, in an attempt to replicate normal ambulation in

the amputee.
The prosthetic heel is the primary area of impact
loading in the prosthesis. As the foot contacts the ground,
the heel is loaded in compression and unloaded slowly as
the amputee moves into mid-stance and the keel loading
begins (Figure 6). In most prostheses, the heel consists
of a compressible foam material that simulates controlled
plantarflexion as it compresses and brings the keel into
contact with the ground. The foam heel uses a viscoelas-
tic material that dissipates energy as it compresses and
expands. Other types of heels include the heel spring
found in the Flex-Foot prosthetic system. The spring acts Figure 6.
Heel compression of a prosthesis during loading response.
like the compressible foam, but with much greater
energy-storage and energy-return capability, initially
compressing and then slowly releasing energy as the foot until the amputee moves forward through stance phase
moves into mid-stance. Thus, for this design, the heel is and begins unloading the prosthesis. As the foot is
an important energy-storage and energy-return part of the unloaded in terminal stance, the keel spring returns a por-
prosthesis. As the stiffness of the heel increases, the dura- tion of the stored energy and assists in propelling the
tion of impact absorption decreases and less energy is limb forward into preswing (Figure 7).
dissipated. The remaining energy is then passed on to the Energy Measurement. Energy measurement and
more proximal sites, such as the socket-residual limb analysis include five interrelated concepts that are used to
interface, or to the musculoskeletal system itself. There- describe the energy performance of a prosthesis: energy
fore, the higher efficiency of the spring-heel comes at a storage, energy return, total energy, dissipated energy,
cost of increased impact absorption by the musculoskele-
and efficiency. These concepts are most easily under-
tal system.
stood by examining the ankle power-time curve gener-
Once the body passes over the foot in mid-stance, ated during kinetic gait analysis (Figure 8). The areas of
loading of the keel begins. In the case of the SACH foot, stored and returned energy are identified as the integrated
the rigid wooden keel deforms minimally to store energy, values of the power-time curve, areas A and B for the
though the soft foam cosmesis compresses and a larger
heel and areas C and D for the keel. The last two con-
amount of energy is dissipated there as stance continues
cepts, dissipated energy (area A minus area B for the
(26). However, in an ESPF with a flexible keel, the keel
heel; area C minus area D for the keel), and efficiency
begins to compress and energy is stored as the foot
moves into dorsiflexion. As the tibial advancement
occurs, the keel spring is compressed and energy stored

Figure 5.
Motion and net muscle action of the foot-ankle complex in walking Figure 7.
gait. Keel loading of a prosthesis during terminal stance.
6

Journal of Rehabilitation Research and Development Vol. 39 No. 1 2002

(B/A for the heel; D/C for the keel) are simply functions dissipates energy during terminal stance requires addi-
of these variables. Since both the heel and the keel can tional energy generation by the amputee’s musculoskele-
store and release energy, their performance features must tal system to achieve the same propulsion, because this
be separated as done here, though in many literature energy is not conserved. With this evaluation method, in
reports, they are not. There are various methods used to terminal stance the Quantum foot dissipated less energy
calculate or measure these variables (as discussed in the (was better) than the SACH, while the Dynamic foot dis-
following paragraphs), but the principles remain the same sipated more energy (was worse) than the SACH (28).
for any method used. Thus, though the Dynamic foot is an energy-storage-and
The various methods developed to measure energy return device and the SACH foot is not, the Dynamic foot
storage capacity of a prosthesis are often used to classify rated less favorably in this energy analysis.
or categorize these prostheses into functional groups. Kinetic Analysis. Kinetic methods are typically the
Unfortunately, the classification systems currently used most common method for evaluating the energy-storage-
do not always agree, and a single foot can be placed in and return capabilities of a prosthetic device. Most
entirely different categories, depending on the analysis motion analysis software packages automatically gener-
method used. Each of the reviewed methods measures or ate the joint powers from the collected kinetic and kine-
calculates one or more of the energy concepts listed pre- matic data. Integration of the joint power (ankle moment
viously. The four primary methods of energy analysis of times angular acceleration) versus time curve can be used
prosthetic feet include functional, mechanical, kinetic, to determine the energy absorbed and released by the
and mathematical analyses. device. The total energy is calculated as the sum of areas
Functional Analysis. The easiest technique that can under the ankle power-time curve (the sum of areas A, B,
be used to characterize or classify the energy characteris- C, and D in Figure 8). Using a total energy calculation,
tics of prosthetic devices is a functional method. Such Ehara grouped the STEN, SACH, Quantum, and Seattle
techniques use a simple performance test with little com- LiteFoot as “low energy”; the Dynamic, Carbon Copy II,
putational analysis. One such method, used by Michael, Seattle, and SAFE as “intermediate energy”; and the
involved attaching each device to a pogo stick and per- SAFE II and Flex-Walk as “high energy” (29,30). Thus,
forming a hopping experiment on each foot (27). The the analysis conducted by Ehara ranks the Dynamic foot
mean maximum vertical displacement of the pogo stick higher than the SACH, yet Van Jaarsveld reverses this
was measured (ten trials), a feature most closely related
ranking (28–30). While Michael ranks the SAFE foot as
to return energy of the keel (area D in Figure 8). With
this criterion, the feet were ranked in order of displace-
ment produced during the hop: SACH, SAFE, STEN,
Carbon Copy II, Seattle, and Flex-Foot (from lesser to
greater displacement). This same ranking corresponded
to subjective clinical evaluations made by the research-
ers. This method does not consider dissipated energy. A
device that requires much energy to deform and has a low
efficiency (but still returns a large amount of energy
compared to similar devices) might still be ranked highly,
although functionally it might be very difficult for an
amputee to use.
Mechanical Analysis. Mechanical analyses are used
to determine the energy characteristics of the prosthesis
in a method similar to that used for standard engineering
materials. The prosthesis is loaded in a mechanical press
(e.g., an Instron® Testing Machine) while force and
deformation are recorded. Hysteresis is proportional to
the dissipated energy, calculated as area C minus area D, Figure 8.
Representative ankle power-time curve.
the energy “lost” during gait (Figure 8). A prosthesis that
 7

HAFNER et al. Transtibial prosthetic devices: Energy concepts

the lowest performing foot (nearly equivalent to the ciency) or SACH (49.5 percent of normal total energy;
SACH, which fractured during testing), Ehara places the 31.0 percent efficiency). Czerniecki’s results were only
SAFE as one of the highest performing feet. Similarly, calculated for the keel section of the foot; performance of
Michael’s test placed the STEN foot as a moderate per- the heel section was not included. However, Czerniecki’s
former, while Ehara ranked it as the very lowest energy- rankings of the Flex-Foot better than the Seattle and the
storing foot, below even the “conventional” SACH foot Seattle better than the SACH are consistent with the
(27,29,30). Clearly, comparison among these types of ranking from Ehara (29,30).
energy transfer analyses results in confusion. Mathematical Analysis. One research group devel-
The total energy, by definition, incorporates both the oped an alternative method for analysis of the energy-
stored and returned energy and therefore might be a bet- storage-and-return characteristics of prostheses. The
ter measure of performance than either alone. However, instantaneous net power was calculated as the sum of the
high total energy derived through the cost of high stored translational (force times velocity) and rotational
energy might not be beneficial to the amputee. As the (moment times angular velocity) joint power components
conservation of energy dictates, large stored energy can throughout gait (33,34). The energy stored and returned
only be accomplished through an energy loss in the in the prosthesis was calculated as the time integral of the
amputee-prosthesis system. Using significant amounts of net power flowing into and out of a fixed point on the
energy from the musculoskeletal system to produce large prosthesis. The method augments the kinematics methods
amounts of energy in the prosthesis might be metaboli- (inverse dynamics model) in order to evaluate the energy
cally detrimental for the amputee or could negatively stored and released in the heel/keel springs, as well as
affect hip or knee wear. The question yet remains as to that stored and released by the cosmesis material. Prince
what amount of energy storage in the prosthesis is ideal. (33) demonstrated that the heel stored and returned a sig-
Further, the total energy does not differentiate between nificant portion of the total stored and returned energy in
the heel and keel sections of the foot. Thus, engineers the foot (46.6 percent stored energy and 19.3 percent
redesigning a foot would have little information on where returned energy for the Flex-Foot, 67.5 percent and
to concentrate design enhancements if only the total 55.7 percent for the SACH foot, and 60.2 percent and
energy, as opposed to the keel energy and heel energy, 50.0 percent for the Seattle Foot, respectively). Thus the
were given. Though Ehara’s methods did allow separa- contribution of the heel-to-energy transfer was signifi-
tion of heel and keel energies in analysis, the separation cant. Independent analysis of the heel portion of the foot
was not included in the measure for prosthesis ranking should thus be part of an energy analysis.
(29,30).
Others have used similar methods to calculate many Nomenclature and Functional Classification
of the energy variables discussed (12,31,32). Czerniecki
used the joint power method to analyze the total work Confusion in Literature
and efficiency of running amputees using the SACH, The limitations in the completeness of energy charac-
Seattle, and Flex-Foot (12). Efficiency is the ratio of the terization of prosthetic feet, particularly early on, led to
returned energy to the stored energy (B/A for the heel an inaccurate nomenclature. Two terms have been used
and D/C for the keel in Figure 8). A device with a large to describe prosthetic feet that can store and return
dissipated energy would therefore have a relatively low energy, ESPF and “dynamic elastic response” (DER)
efficiency. Efficiency is usually obtained at the cost of an foot. The term “ESPF” was first adopted in the late 1980s
increase in stiffness of the spring material. Since effi- to differentiate feet with a flexible keel design from those
ciency is a calculated ratio, a high efficiency may be without (typically the SACH foot) (6,27). However, the
obtained at any magnitude of energy storage or return so term ESPF is an inadequate description, because it con-
long as the ratio approaches unity. Czerniecki avoids this siders only areas A and C in the ankle power-time curve
limitation by reporting both the efficiency and the total (Figure 8). The term “DER” foot or “dynamic response”
energy. In that study, the Flex-Foot produced higher total foot was developed in the early 1990s (2). DER, while
energy (70.0 percent of the normal control) and spring signifying perhaps the flexing action of the prosthetic
efficiency (84.0 percent efficiency) than either the Seattle foot, does not describe the loss of energy, but merely the
(63.0 percent of normal total energy; 52.0 percent effi- transfer of potential energy to kinetic energy. This
8

Journal of Rehabilitation Research and Development Vol. 39 No. 1 2002

terminology disregards the dissipated energy in the heel

(area B minus area A, Figure 8) and in the keel (area D
minus area C, Figure 8). Further, the term “elastic” sug-
gests equivalent areas of energy storage and return in the
prosthesis. In order to be accurate, a terminology should
address all four areas of energy storage or return repre-
sented in the power-time curve (areas A, B, C, and D in
Figure 8).
Figure 9.
A Revised Terminology Typical energy determinants used for analysis of prosthetic devices.
In order to adequately describe the function and per-
formance of lower-limb prosthetics, a modified conven-
sports. Adopting a universal and technically accurate ter-
tion for the description of these devices is proposed.
minology will lead to a greater general understanding and
Extending from the original ESPF terminology, the term
consistency in measurement techniques of prosthetics
energy-storage and energy-return prosthetic foot is sug-
technology.
gested. This term signifies not only that the device can
store, but also can return energy to the amputee during
gait.
DISCUSSION
It is further suggested that to accurately classify an
energy-storage-and return device, at least four attributes
are required: two for the heel and two for the keel. Two This paper examines the modern “energy-storing”
of each of the following must be reported for both the prosthetic device through a presentation of energy con-
heel and keel: either energy storage or energy return and cepts, a recommendation for a revised and technically
either energy efficiency, energy dissipation, or total consistent nomenclature, and a new technique for analy-
energy. Energy storage and energy efficiency are pre- sis and categorization based on performance for energy-
ferred. The energy efficiency of the device corresponds storage-and return prostheses. A review and critical anal-
to the response, while the energy storage corresponds to ysis of the relevant literature reveals key issues to be
accommodation (7). From these two parameters, the addressed in order to adequately understand and investi-
three additional parameters may be calculated easily for gate amputee performance and the fundamental design of
each section of the foot (Figure 9). advanced energy-storage-and return prosthetic devices.
With the use of this convention, a single quantity can Knowledge of the mechanisms of energy transfer in
be used to describe both the heel and keel behavior. The the energy-storing foot is necessary to understand the
heel-keel (HK) system is derived from the efficiency and
energy storage capacity of each section of the foot (Fig-
ure 10). A heel with low energy storage (accommoda-
tion) is not clinically functional to amputees; therefore
categories H3 and H4 have been eliminated from the
metric. With the use of this metric, there are eight
remaining possible device descriptions corresponding to
the characteristics of accommodation and response of the
heel and keel (Table). Of the eight, only five clinically
functional combinations remain. This improved HK cate-
gorization process will provide clarification of usage in
the literature and convey the proper action and applica-
tion of these devices. For example, a high-energy
(accommodation), high-efficiency (response) heel and a Figure 10.
high-energy, high-efficiency keel (H1K1) might be pre- Proposed heel-keel (HK) functional energy-storage-and return
scribed for a very active patient who runs and plays prosthesis evaluation system.
 9

HAFNER et al. Transtibial prosthetic devices: Energy concepts

Table.
HK categorization chart and associated activity level.
Walking Sports Other
Heel Heel Keel Keel
Categori- Speed Activity Properties
Accommodation Response Accommodation Response
zation Uneven Shock Increased
(Energy Stg) (Efficiency) (Energy Stg) (Efficiency) Slow Moderate High Running
Terrain Absorption Simulated ROM
H1K1 High High High High — • • • — — •
H2K1 High Low High High — • • • — • •
H2K3 High Low Low High • • — — — • —
H2K2 High Low High Low • • — — • • •
H2K4 High Low Low Low • • — — — • —
H1K2 High High High Low — — — — — — —
H1K3 High High Low High — — — — — — —
H1K4 High High Low High — — — — — — —
= Characteristics not suitable for a clinically functional prosthetic foot.

performance and application of advanced prosthetic tion provides a total of five functional foot categories and
devices. Although implicitly understood, the literature is tabulated according to functional capacity for future
often fails to designate and differentiate the two energy- use. This metric is designed not to be a complete replace-
storing components of a prosthetic foot—the heel and the ment, but an augmentation to existing clinical evaluation
keel. Since the major function of an energy-storage-and systems, adding a measure of performance that is both
return device is to propel the amputee during gait, the functional and technically precise. More extensive
keel performance is often the focus of analysis while the research and clinical evaluation are now required to com-
heel is often overlooked. One reason for this oversight plete and augment the proposed HK classification sys-
might be that the heel’s returned energy often is not uti- tem. While classical engineering methods and gait
lized as an input energy for the keel, but rather simply analysis can be used to analyze the accommodation and
dissipated. A future area of research might be examining response of these devices, clinical input is required to
energy transfer between these two areas of the foot. The determine the functional application and recommended
question of whether the energy stored in the foot during usage of each level of the scale.
loading response (heel energy storage) can be transferred
Finally, one must understand that the tools of engi-
into the terminal stance energy release (keel energy
neering methods and gait analysis, while helpful, are only
release) or whether it can only be dissipated needs to be
answered. If prosthetic designs can be modified to more a part of the whole design of a successful rehabilitation.
efficiently utilize input energy rather than to dissipate it, A performance analysis of the prosthesis may help pre-
then amputee performance might be enhanced. dict the behavior of the device when used by the individ-
A revised terminology, dubbed energy-storage-and ual, but is not entirely sufficient to analyze the
return, is proposed to accurately describe the function performance of an amputee. Future research must con-
and performance of advanced prosthetic devices. While centrate not on analyzing which devices work, but on
this terminology does not necessarily distinguish the analyzing why the devices that do work are successful for
types of prostheses by definition alone, the measured the particular amputee. A vital component to analyzing
magnitudes of the energy transfer parameters can be used prosthesis energy transfer should now be to understand
to categorize the devices by functional performance. This how much energy absorption and release is appropriate
categorization will then provide a clinical tool for align- for the individual, as well as understand how that transfer
ing amputee activity level and device performance in a of energy affects the individual. Questions such as, “Is it
way currently only indirectly discussed in the literature. better for the amputee to absorb more or less energy in
A proposed classification system is suggested in order to the prosthetic limb than the sound limb (or that of a nor-
rank and categorize energy-storage-and return and con- mal limb)?” must be answered to further develop
ventional feet based upon the accommodation and advanced prosthetic designs. While the performance of
response of each section of the foot. This HK classifica- the prosthesis will always be a vital component of
10

Journal of Rehabilitation Research and Development Vol. 39 No. 1 2002

prosthetic design, the ultimate goal will always be the 10. Winter DA. Energy generation and absorption at the ankle
optimum performance and health of the amputee. and knee during fast, natural, and slow cadences. Clin
Toward this end, analysis of energy transfer mecha- Orthop 1983;175:147–54.
nisms in an energy-storage-and return prosthesis has 11. Bresler B, Frankel JP. The forces and moments in the leg
during level walking. Trans ASME 1950;72:27–36.
been examined and the fundamental characteristics
12. Czerniecki JM, Gitter A, Munro C. Joint moment and mus-
explained. A revised nomenclature and a system for cate-
cle power output characteristics of below knee amputees
gorization based upon functional performance have been during running: the influence of energy storing prosthetic
suggested for energy-storage-and return prosthetic feet. J Biomech 1991;24(1):63–75.
devices. Further input is required from researchers and 13. Nielsen DH, Shurr DG, Golden JC, Meier K. Comparison
clinicians at large to expand upon the ideas presented of energy cost and gait efficiency during ambulation in
here and to further evaluate both amputee and energy- below-knee amputees using different prosthetic feet—a
storage-and return prosthesis performance. By maintain- preliminary report. J Prosthet Orthot 1989;1(1):24–31.
ing an understanding of the energy principles of the pros- 14. Torburn L, Perry J, Ayyappa E, Shanfield SL. Below-knee
thesis, a consistent and technically accurate amputee gait with dynamic elastic response prosthetic feet:
nomenclature, and a method to categorize the energy- a pilot study. J Rehabil Res Dev 1990;27(4):369–84.
storage-and return devices according to functional per- 15. Macfarlane PA, Nielsen DH, Shurr DG, Meier K. Gait
formance, we achieve a better position from which to comparisons for below-knee amputees using a Flex-Foot
provide greater care for the amputee and an increased versus a conventional prosthetic foot. J Prosthet Orthot
ability to improve upon prosthetic designs. 1991;3(4):150–61.
16. Colborne GR, Naumann S, Longmuir PE, Berbrayer D.
Analysis of mechanical and metabolic factors in the gait of
congenital below knee amputees: A comparison of the
REFERENCES SACH and Seattle feet. Am J Phys Med Rehabil
1992;71(5):272–8.
1. Postema K, Hermens HJ, de Vries J, Koopman HF, Eisma 17. Lehmann JF, Price R, Boswell-Bessette S, Dralle A, Ques-
WH. Energy storage and release of prosthetic feet. Part 2: tad K. Comprehensive analysis of dynamic elastic response
Subjective ratings of 2 energy storing and 2 conventional feet: Seattle Ankle/Lite Foot versus SACH foot. Arch Phys
feet, user choice of foot and deciding factor. Prosthet Med Rehabil 1993;74(8):853–61.
Orthot Int 1997;21(1): 28–34. 18. Lehman JF, Price R, Boswell-Bessette S, Dralle A, Questad K,
2. Menard MR, McBride ME, Sanderson DJ, Murray DD. deLateur BJ. Comprehensive analysis of energy storing pros-
Comparative biomechanical analysis of energy-storing thetic feet: Flex Foot and Seattle Foot versus standard SACH
prosthetic feet. Arch Phys Med Rehabil 1992;73(5):451–8. Foot. Arch Phys Med Rehabil 1993;74(11):1225–31.
3. Murray DD, Hartvikson WJ, Anton H, Hommonay E, Rus- 19. Powers CM, Torburn L, Perry J, Ayyappa E. Influence of
sell N. With a spring in one’s step. Clin Prosthet Orthot prosthetic foot design on sound limb loading in adults with
1988;12(3):128–135. unilateral below-knee amputations. Arch Phys Med Reha-
bil 1994;75(7):825–9.
4. Alaranta H, Kinnunen A, Karkkainen M, Pohjolainen T,
20. Schneider K, Hart T, Zernicke RF, Setoguchi Y, Oppen-
Heliovaara M. Practical benefits of Flex-Foot in below-
heim W. Dynamics of below-knee child amputee gait:
knee amputees. J Prosthet Orthot 1991;3(4):179–81.
SACH foot versus Flex Foot. J Biomech 1993;
5. Edelstein JE. Prosthetic feet. State of the Art. Phys Ther 26(10):1191–204.
1988;68(12):1874–81.
21. Wagner J, Sienko S, Supan T, Barth D. Motion analysis of
6. Wing DC, Hittenberger DA. Energy-storing prosthetic SACH vs. Flex-Foot in moderately active below-knee
feet. Arch Phys Med Rehabil 1989;70(4):330–5. amputees. Clin Prosthet Orthot 1987;11(1):55–62.
7. Romo HD. Specialized prostheses for activities. Clin 22. Prince F, Allard P, Therrien RG, McFadyen BJ. Running
Orthop 1999;361:63–70. gait impulse asymmetries in below-knee amputees. Pros-
8. Miller LA, Childress DS. Analysis of a vertical compliance thet Orthot Int 1992;16(1):19–24.
prosthetic foot. J Rehabil Res Dev 1997;34(1):52–57. 23. Perry J, Shanfield S. Efficiency of dynamic elastic
9. Hsu MJ, Nielsen DH, Yack HJ, Shurr DG. Physiological response prosthetic feet. J Rehabil Res Dev 1993;
measurements of walking and running in people with tran- 30(1):137–43.
stibial amputations with 3 different prostheses. J Orthop 24. Snyder RD, Powers CM, Fontaine C, Perry J. The effect of
Sports Phys Ther 1999;29(9):526–33. five prosthetic feet on the gait and loading of the sound
 11

HAFNER et al. Transtibial prosthetic devices: Energy concepts

limb in dysvascular below-knee amputees. J Rehabil Res amputee walking. Am J Phys Med Rehabil 1991;70(3):
Dev 1995;32(4):309–15. 142–8.
25. Barth DG, Schumacher L, Thomas SS. Gait analysis and 32. Barr AE, Siegel KL, Danoff JV, McGarvey CL, Tomasko
energy cost of below-knee amputees wearing six different A, Sable I, Stanhope SJ. Biomechanical comparison of the
prosthetic feet. J Prosthet Orthot 1992;4(2):63–75.
energy-storing capabilities of SACH and Carbon Copy II
26. Winter DA, Sienko SE. Biomechanics of below-knee
prosthetic feet during the stance phase of gait in a person
amputee gait. J Biomech 1988;21(5):361–7.
with below-knee amputation. Phys Ther 1992;72(5):344–54.
27. Michael J. Energy storing feet: a clinical comparison. Clin
Prosthet Orthot 1987;11(3):154–68. 33. Prince F, Winter DA, Sjonnesen G, Wheeldon RK. A new
28. Van Jaarsveld HW, Grootenboer HJ, de Vries J, Koopman technique for the calculation of the energy stored, dissi-
HF. Stiffness and hysteresis properties of some prosthetic pated, and recovered in different ankle-foot prostheses.
feet. Prosthet Orthot Int 1990;14(3):117–24. IEEE Trans Rehab Eng 1994;2(4):247–55.
29. Ehara Y, Beppu M. Nomura S, Kunmi Y, Takakashi S. 34. Prince F, Winter DA, Sjonnensen G, Powell C, Wheeldon
Energy analysis of energy-storing prosthetic feet. Proceed- RK. Mechanical efficiency during gait of adults with tran-
ings of the International Symposium on Gait Analysis: stibial amputation: a pilot study comparing the SACH,
1990 Feb; Berlin, Germany. p. 298–307.
Seattle, and Golden-Ankle prosthetic feet. J Rehabil Res
30. Ehara Y, Beppu M, Nomura S, Kunmi Y, Takakashi S.
Dev 1998;35(2):177–85.
Energy storing property of so-called energy-storing pros-
thetic feet. Arch Phys Med Rehabil 1993;74(1):68–72.
31. Gitter A, Czerniecki JM, DeGroot, DM. Biomechanical Submitted for publication January 5, 2001. Accepted in
analysis of the influence of prosthetic feet on below-knee revised form March 13, 2001.