Four Bar Linkage Knee Analysis

by Michael P. Greene, B.S., M.E., C.P.O.

INTRODUCTION learn comparative methods of evaluating

the efficiency of a particular four bar de­
Modern prosthetists have a wide selec­ sign in attaining its specific mechanical or
tion of prosthetic knees to fulfill many in­ cosmetic goals. This skill is extremely im­
dividual specifications. The names "fric­ portant since each four bar design is
tion," "safety," "lock," "hydraulic," etc. unique in its operation. Specifically, each
quickly recall particular classes of single four bar knee has been designed to en­
axis knees. For these single axis knees, the hance individual characteristics such as
name (friction, safety, etc.) simply states a safety, cosmesis, energy conservation
unique feature which defines the major and/or swing phase motion.
mechanical advantage of that class of
knees.
Polycentric knees, however, may pre­
sent the prosthetist with confusion. This DEFINITION OF TERMS
confusion results from the fact that the 1. Translation or translational motion is
term "polycentric" does not define any the movement of a machine element along
specific function. Secondly, these knees a straight line.
require more than a simple knowledge of 2. Rotation or rotational motion is the
mechanics to fully understand their func­ movement of one element of a mechanism
tions. about a pivot point.
This paper will examine one category of 3. Center of Rotation is the point about
polycentric knees which are known as which rotational motion occurs. This may
"four bar linkages." Simple methods for be an actual mechanical pivot point on the
evaluating these knees will be presented. mechanism or a purely hypothetical point
These evaluating methods will enable which may or may not actually be on the
the prosthetist to determine the major mechanism.
mechanical or cosmetic advantage of most 4. Single Axis Knee—Any knee in which
four bar designs. The prosthetist will also the shin moves in pure rotation about a
constant center of rotation located at the be a pair of parallel links acting together.
knee bolt. However, for mechanical purposes these
5. Polycentric Knee—Any knee whose pairs are considered as single links.)
design allows the shin to move in a com­ Fig. 1A is a typical four bar linkage
bination of rotational and translational knee. The thigh is considered as a link or
motion. At any given instant of time, this bar joining points B and E. This link is
combination can be mechanically de­ defined BE. The shin is considered as a
scribed as a purely rotational motion link joining points C and D. This link is
about a constantly changing center or ro­ called CD. Link BC and ED join the shin to
tation known as the instantaneous center the thigh. Together, all four links join at
of rotation. four points to complete the four bar link­
6. Instantaneous Center of Rotation (or age. Fig. I B is a kinematic schematic rep­
Instant Center)—The point about which a resentation of the knee seen in Fig. 1A
particular element (shin) may be assumed which shows this typical link arrange­
to be moving in pure rotation at any given ment.
instant of motion being analyzed. For a
single axis knee this will be a constant
point at the knee bolt center. For a
polycentric knee this will be a theoretical STABILITY IN STANCE
point in the plane of motion (sagittal PHASE OF A FOUR BAR
plane).
7. Four Bar Linkage Knee—A specific
LINKAGE KNEE
class of polycentric knees. The knees are Alpha (a) Stability—At this point it
characterized by four elements joined at is assumed that the reader understands
four separate points. The four elements the basic theory of the T.K.A.(Trochanter-Knee-Ankle)line a
include the thigh, shin and two links. T.K.A. alignment method of simple single
(Note: In actual practice, a single link may
axis knee mechanisms. In this method the the equivalent single axis knee will also
knee is made more stable (safer) by mov­ be different for each position of flexion.
ing the knee center posterior to the T.A. Therefore, care must be taken to analyze
(Trochanter-Ankle) line. Conversely, the four bar mechanism at the exact angu­
moving the knee center anterior to the lar position which is in question.
T.A. line decreases weight bearing sta­ A simple method of estimating the in­
bility. stantaneous center of rotation of an actual
Stability of a four bar knee system is four bar knee mechanism would be to lay
also determined by using the T.K.A. two straightedges along the links and note
theory. The knee center becomes the the point of intersection. A third straight­
theoretical "instantaneous center of rota­ edge could be aligned with the trochanter
tion" in this case. This point must be de­ and ankle center to simulate the T.A. line.
termined for each position of the knee Stability of the system is estimated by
which is in question. measuring the distance from the T.A. line
For static (bench) alignment purposes, to the instant center. For the sake of this
the accepted knee position is that of full discussion, this distance will be defined
extension. With the knee fully extended as "a" (alpha). A positive a value is de­
the instantaneous center for rotation is fined as a knee center which is posterior to
determined by drawing a line through the T.A. line. This is a stable or "positive a
each of the two links joining the shin to stability" condition. A negative a value
the thigh (see Fig. 1A). The instantaneous indicates an unstable system with the
center of rotation (point O) is the point knee center anterior to the T.A. line.
where these two lines intersect. The sta­ At this point it is interesting to compare
bility of the system is determined by not­ a prosthesis with a single axis knee to the
ing the position of this instant center in four bar knee prosthesis seen in Fig. 1A.
relation to the T.A. line. As in the single The single axis knee has an a = 0 value at
axis knee, the center of rotation must be
posterior to the T.A. line to be considered
as a stable weight bearing system.
At this point the reader's understanding
of the "instantaneous center of rotation"
and of four bar knee motion may be un­
clear; this confusion can be eliminated if
one understands that a four bar knee is
mechanically equivalent to a particular
hypothetical single axis knee at any in­
stant of motion being analyzed. This hy­
pothetical knee has its knee bolt located at
the instant center of the equivalent four
bar knee. F i g . l C gives the single axis
equivalent of the four bar knee depicted in
Fig. 1A (at the full extension position
only). Therefore, the motion and mech­
anical reaction of the four bar knee in
Fig. 1A is precisely identical to that of the
single axis knee seen in Fig. 1C at this
position of extension. Often it is easier to
understand the reaction of the four bar if
one visualizes this instantaneous single
axis equivalent rather than the actual four
bar mechanism.
Since the instant center of a four bar is
changing through each position of flexion, F i g . L-C
full extension. As it begins to flex, a be­ nent of load applied at the knee bolt by the
comes negative and progressively more thigh section. The force E is the force
unstable as flexion continues. The special applied to extend the knee mechanism.
four bar knee in Fig. 1A has a positive a This force is also applied by the thigh at
value at full extension. As flexion begins, the knee bolt. Forces R and R are the
v h

the value becomes smaller but it remains vertical and horizontal components of the
positive for the first few degrees of flexion. floor reaction force. To analyze this situa­
Obviously, this knee was designed to tion, moments are summed to equal zero
have enchanced stance stability and about the point " f " to yield the equation:
therefore could accurately be called a "four
bar safety knee."
It is noted that if the knee center is raised,
Beta (B) Stability—A second and the value of "y" and of L will remain un­
unique condition affecting knee stability changed. However, the value of " h " will
exists with all four bar knee mechanisms. increase and for the above equation to
Referring to Fig. 1A, it is noted that the balance; the value of E will proportion­
instantaneous center of rotation is super­ ately decrease. This simply means that the
ior to the level of the mechanical (or cos­ moment tending to cause knee buckling is
metic) knee center (point K ) . With this
c reduced and therefore the patient uses less
prosthetic knee the patient gains a force, E, to hold the knee in extension.
mechanical advantage over a typical single The second way in which knee stability
axis knee. This mechanical advantage is is increased by raising the knee center is
gained in two ways as a result of raising demonstrated in Fig. 2B. This represents a
the instant center. typical above knee prosthetic thigh. Force
Fig. 2A is a free body diagram of a typi­ W and I are the loads applied to socket by
cal above knee prosthetic shin shortly after the patient, (note: W and I are assumed to
heel strike. The force L is the axial compo­ act on a point along the T.K.A. for this
analysis. L ' is the axial component of selecting a particular four bar mechanism
reaction force applied by the shin at the when "safety" or "stability" are primary
knee bolt (L" = - L ) . E" is the force applied concerns. It is interesting to note that both
by the shin tendng to buckle the knee (E" a and B stability are permanently built
= — E). H is the extension force applied by into a prosthesis and do not require
the residual limb to hold the knee in ex­ maintenance or adjustment as is typical o f
tension. x is the effective lever arm of the
2 single axis safety knees, a and B stability
residual limb. To analyze this situation, are also independent of any extension
moments are summed about the point " t " aids, hydraulic mechanism, etc.
to equal zero: (WARNING: a and B stability are fea­
tures of only certain four bar mechanisms
which were originally designed for stabil­
It is noted that if the knee center is raised, ity. Some four bar mechanisms may be de­
the value of x would remain constant.
2 signed for cosmetic or swing phase
This condition would also decrease the characteristics and therefore may have
value of E (reduce buckling force as seen poor values of a and B stability.)
above) and thus reduce the values of E '
and H proportionately. It is also observed
that X j would decreasejn value creating a SHORTENING OF A FOUR
second way in which H would be propor­
tionately decreased. This second advan­ BAR KNEE PROSTHESIS
tage can also be described as increased DURING SWING PHASE
leverage for the residual limb.
In summary, raising the knee center re­ With the standard single axis knee
duced the knee buckling moment and in­ prosthesis a typical problem encoun­
creases the patients leverage advantage in tered is that of foot to floor clearance during
controlling that moment. With single axis swing phase. It is sometimes necessary to
knees these advantages would only be avail­
able by sacrificing the cosmetic appear­
ance of bending at the anatomical knee
center. This is not the case with a four bar
knee mechanism. The four bar knee can
give the cosmetic appearance of bending
at the proper anatomical height while pro­
viding the added stability of a proximal
instantaneous knee center. Fig. 1A depicts
a typical four bar knee prosthesis and its
anatomic (or cosmetic) knee center, Kc • /3
(beta) is the vertical difference between
the anatomical knee height and the in­
stantaneous knee center at full extension.
The B value (or "B stability") gives a rela­
tive value of stability for comparing four
bar mechanisms to each other and to
single axis knees. B is measured positive if
the instantaneous knee center is above the
anatomic knee center, and conversely
negative if this instant center is lower than
the anatomical center.
The simple method outlined previously
for determining the instant center will also
yield stability. By determining these
values the prosthetist now has a guage for Fig. 3-A.
shorten the prosthesis excessively to pro­ ACCELERATION/
vide floor clearance during swing phase.
Certain four bar knees, however, actually DECELERATION OF A
shorten as they pass from full extension to FOUR BAR LINKAGE KNEE
flexion. This feature allows fabrication of a
"full length" prosthesis which automati­ DURING SWING PHASE
cally "shortens" during swing phase, Precise kinematic and dynamic studies
similar to the actual human knee joint. of four bar knee units can be extremely
Fig. 3A depicts a four bar knee at full complex. Therefore, this paper will not at­
extension. The thigh length is "A " and lt
tempt to analyze the complex motion of
the shin length is " B ^ " The overall pros­ these mechanisms by any quantitative
thesis length is " O " as follows: means. In lieu of a detailed analysis, a gen­
eral qualitative examination will be pre­
sented.
In Fig. 3B the mechanism is in the 65° flex­ Basic single axis knees with mechanical
ion position,* which is generally accepted friction and spring assisted extension are
as the "mid swing" position. The value of essentially "linear" in their response dur­
A + B or C has now decreased and
2 2 2 ing swing phase. The term "linear" applies
therefore results in additional foot to floor a "constant" or "constant rate of change" of
clearance. The amount of overall shorten­ some property of the system. The mechani­
ing is defined as the " L " value: cal friction is constant regardless of knee
L = d - C (at 65° flexion)
2
position or velocity. The spring assisted
extension assist constantly increases (ap­
L values for common four bar knee proximately) as knee flexion increases. The
mechanisms are given in Table 1. extension assist is also independent of
knee velocity. Both of these features are
adjustable to allow "tuning" or the swing
phase characteristics of "heel rise" at "toe
off" and impact at full extension.
Often it is impossible to suit a particular
patient's gait pattern by tuning a basic
single axis knee. Adjustment of friction or
extension may cure one problem only to
create another. Although both heel rise
and terminal impact may finally b e ad­
justed to prosthetic tolerances, the result
may be a system that requires excessive
effort by the patient. In this case, the pa­
tient often insists that the system be ad­
justed to suit his requirements for ease of
flexion at the sacrifice of smooth operation.
Four bar knees are nonlinear in their
operation. As the position of the shin
changes, acceleration (deceleration) vary
relative to position. This variance can be
nonlinear depending on the design of the
four bar mechanism. Therefore, it is possi­
ble to design a knee with motion charac­
teristics similar to normal human knee
motion. For example, certain four bar knee
Fig. 3 - B . designs have built-in terminal deceleration
which requires no use of mechanical fric­
* Page 178 Northwestern University A b o v e Knee Prosthetic Manual tion or other devices.
 To understand the acceleration-decel­ benefit of a four bar prosthesis which
eration of a four bar mechanism the shin simulated the motion of the actual human
can be compared to the pendulum of a knee joint. This advantage also gives the
pendulum clock. By lowering the weight unilateral above knee amputee the visual
on the pendulum, the effective pendulum appearance of legs with matching knee
moment arm is lengthened. This adjust­ heights when sitting.
ment slows the pendulum movement. For tall amputees, an excessively long
Raising the weight conversely increases shin can cause clearance problems when
the speed of the pendulum. In the four bar sitting at desks or tables. In addition, when
knee the "pendulum moment arm" is in­ sitting on low chairs the tall amputee is
creased as the instant center movesproximally during flexion.position
forced into an uncomfortable This action
of slows th
shin movement and causes the decelera­ excessive hip flexion. The four bar knee
tion phenomenon. Conversely, as the in­ reduce both of these problems by the
stant center moves distally, the shin accel­ shortening action of the shin when sitting.
erates. The " L " value was defined above at 65°
knee flexion to provide a comparative
As stated above, the precise quantitative method of analyzing shortening of a pros­
analysis of a four bar motion is very dif­ thesis. If the same calculation is made at 90°
ficult. However, the prosthetist can ob­ of knee flexion, the value obatined would
serve the operation of these knees and then be the effective shortening of the prosthetic
make certain qualitative judgments re­ shin when sitting. This value is defined as
garding the swing phase characteristics of the "S" value. S values for common four
a particular mechanism. Terminal deceler­ bar knees are listed in Table 1.
ation and response time (from "toe off" to S = C1 - C (at 90° flexion)
2
full extension) are two characteristics (see Fig. 3A and 3B)
which are very easy to observe. These ob­
servations can be made by either manually Special Case-Knee Disarticula­
swinging the knee mechanism or by actual tion—Conventional single axis knees pre­
testing on a patient. sent a particular cosmetic problem when
It should be noted that hydraulic and fitting long above knee or knee disarticu­
pneumatic knee mechanisms are also con­ lation amputations. With these amputa­
sidered "nonlinear" in their operation. tions, it is impossible to fabricate a pros­
However, this nonlinearlity is not the same thesis with a knee center at the anatomical
as that of a four bar mechanism. Hydraulic height unless outside joints are used.
and pneumatic knees respond nonlinearily However, outside joints have no friction
to different velocities of operation. This is not adjustment, are not durable, and increase
the case with four bar mechanisms. Four knee width. The distal end of the socket
bar mechanisms are nonlinear with respect can only be placed wihin 1/2 to 21/4inches
to shin position; not velocity. If a four bar proximal (depending on the particular
mechanism is desired which automatically knee mechanism) to the knee bolt center
adjusts to varied gait speed, that mecha­ when a conventional above knee joint is
nism must incorporate a hydraulic or used. In the case of knee disarticulation
pneumatic unit. this could require lowering the prosthetic
knee center approximately 2 to 4 inches
below the anatomical (cosmetic) knee
ADVANTAGES OF A FOUR center, resulting in an excessively long
BAR KNEE IN THE thigh and short shin components. This
condition is cosmetically unsightly when
SITTING POSITION sitting and causes clearance difficulties
General case—A sitting advantage of when sitting in confined areas such as the
a four bar knee is the effective shortening rear seat of small automobiles.
of the shin as it passes into flexion. This With certain four bar knee designs it is
feature was noted above as a swimg phase possible to place the distal end of the socket
 Fig. 4 - B .

factor." If the point D is distal to the point

C, the K factor is positive. If the converse is
true, the K factor is negative. A positive K
factor indicates a "cosmetic advantage"
over single axis systems (Note: C is the
"Cosmetic Knee Center").
All single axis knee shin units have
Fig. 4-A. negative K factor values ranging approxi­
mately from minus 1/2 to minus 21/2inches.
Outside joints, however, have a positive K
at a level distal to the cosmetic (anatomical) factor value that can be as large as nec­
knee center. A simple method of quantita­ essary.
tively evaluating this property of "cosmetic Table # 1 lists K factors for the most
advantage" for a four bar knee is presented common four bar linkage knees. Those
in Fig. 4A and Fig. 4B. knees with a positive K factor would give
Fig. 4A is a schematic of an endoskeletal the best cosmetic knee center for knee dis­
four bar knee mechanism that has been articulation amputations. Those knees
designed to have the aforementioned with negative K factors would tend to be
"cosmetic advantage." The point D is the undesirable cosmetically for knee disar­
most distal position at which the socket can ticulation amputations.
possibly be placed along the T.A. line. With Positive K factors and L values are not
the knee mechanism fully extended, the the only property affecting true cosmetic
T.A. line is noted on both the shin and analysis of a four bar knee. Each mecha­
thigh. The point D is also noted. The knee nism must be judged by the individual
mechanism is then flexed 90° as seen in Fig. prosthetist to determine the ease of
4B. The point at which the shin T.A. line finishing or the general appearance of the
and the thigh T.A. line intersected is noted finished prosthesis. Certain four bar
and defined as point " C . " Finally, the dis­ mechanisms may have positive K factors
tance from point C to point D is measured but may be difficult to finish-fabricate with
and this value is defined as " K " or the " K an acceptable cosmetic appearance.
Table # 1 w a s c o m p o s e d using methods w h i c h are graphical. Therefore the values derived are subject to a wider
m a r g i n o f error t h a n p u r e l y calculated v a l u e s . M a n u f a c t u r e r s s h o u l d b e c o n s u l t e d for m o r e precise data.
T h e v a l u e s t a b u l a t e d w e r e m e a s u r e d u s i n g t h e m e t h o d s d e s c r i b e d a b o v e . T h e T . A . l i n e w a s assumed t o b e a
l i n e p a s s i n g t h r o u g h t h e f o o t b o l t a n d t h e c e n t e r o f t h e p y l o n o r s h i n u n i t . T h i s a s s u m p t i o n w a s m a d e s i m p l y to
provide a uniform m e t h o d o f evaluating a n d c o m p a r i n g four bar k n e e designs. This T.A. line should not
necessarily b e u s e d for a l i g n m e n t p u r p o s e s . F o r this p u r p o s e the m a n u f a c t u r e r ' s instructions s h o u l d b e strictly
followed.
T h e P o l y m a t i c a n d P o l y c a d a n c e k n e e s are not in p r o d u c t i o n at this t i m e b u t are i n c l u d e d to s h o w the
uniqueness of each four b a r design.
Certain k n e e s can be tilted in the sagittal plane. This feature allows s o m e adjustment of a a n d B values by
m o v i n g the instant c e n t e r relative to the T.A. line. T h e v a l u e s t a b u l a t e d i n T a b l e # 1 w e r e m e a s u r e d w i t h all
knees in the vertical position (no tilt).
T h e O t t o B o c k 3 R 2 0 k n e e has a n adjustable extension stop w h i c h adjusts all of the values listed. Therefore the
least stable a n d m o s t stable positions a r e listed to s h o w t h e full r a n g e o f a d j u s t m e n t .
ENERGY CONSUMPTION Since mechanical friction is an energy con­
suming phenomenon, this furnished an
WITH FOUR BAR KNEE additional means of energy conservation
MECHANISMS for certain four bar knees.
Finally, the acceleration and deceleration
The amputee consumes energy during of a four bar knee are relative to knee posi­
ambulation through muscular activity. tion. In effect, these properties are per­
This muscular activity develops the forces fectly timed controls occuring only at the
necessary for ambulation. It is the goal of position at which they are required. The
the prosthetist to eliminate unproductive precision and efficiency thus provided can
forces and minimize the productive forces also serve as a source of energy savings.
required of the patient. This results in a
proportional decrease in the energy loss of
the patient during ambulation. CONCLUSION
It was shown above that a and B stability
reduce the force required from the patient Four bar knee mechanisms can provide
to maintain extension during the early part the prosthetist with a selection of knee
of stance phase. This force reduction re­ characteristics which were previously un­
sults in a directly proportional energy available with a single axis knees. The
savings and therefore, a and B give a rela­ prosthetist should, through simple analy­
tive means of evaluating this energy loss. sis of any four bar mechanism, be able to
It was noted that the four bar knee pros­ define the unique qualities or advantages
thesis can shorten as it passes from exten­ of that knee mechanism. With this skill the
sion to flexion. This feature eliminates prosthetist can confidently select a four bar
energy losses due to gait defects such as knee to meet the specific needs of an indi­
"hip hiking," "vaulting," "circumduct­ vidual prosthetic patient.
ing," etc. This feature also eliminates the
need for excessive shortening of the pros­ ACKNOWLEDGMENTS
thesis. The amount of prosthetic shorten­ The author wishes to express appreciation to Barbara Welsch for
ing causes a directly proportional energy editorial and clerical assistance in preparation of this manuscript, to
Steven O h m e r for his excellent graphics, and to William Svetz, C P.O
loss. Moving the patient's mass center up for technical assistance
and down during each full cycle of gait is
the source of this loss. Therefore, the L BIBLIOGRAPHY
value gives a relative means of analyzing Hinkle, R.T., Konematics of Machines, Prentice-Hall, Inc., E n g l e w o o d
the reduction of this particular energy loss. Cliffs, N J 1960.
Muhbauer, K . C . , Statics— an Individual Approach, Addison-Wes­
The special acceleration-deceleration ley— Publishing C o m p a n y , Reading, Mass., 1972
properties of certain four bar mechanisms
also contribute to energy savings. The effi­ Michael P Greene, B . S . M . E . , C.P.O. is currently Branch Manager,
Union Artificial Limb & Brace C o m p a n y , Greensburg, Pennsylvania
cient operation afforded by these knees re­ He was formerly a Product Design Engineer for A.O. Smith Cor­
duces the need for mechanical friction. poration