2. Aim
• To develop a theory of walking that is
biomechanically rigorous and clinically
meaningful.
• A clinically meaningful model is one that
provides insights into how we can help our
patients to walk more easily.
2
3. 3
“ … it is obvious that any improvement -
either in surgical and physiotherapeutic
procedures or in braces and prostheses -
must rest upon an accurate knowledge
of the functional characteristics of the
normal locomotor system.”
Eberhart, Inman and Bresler
Human Limbs and their Substitutes (1954)
4. Motor control
• Biomechanics can tell us why we walk the
way we do
• Motor control tells us how we achieve this
• This lecture, following, my expertise, will
focus on the biomechanics
4
6. Hip flexion
Pelvic rotation
Stance phase
knee flexion Swing phase knee flexion
Compass
gait
Successively smoothing the trajectory of the centre of mass
7. “Translation of the body in a straight line with the
least expenditure of energy may be achieved
mechanically by means of a wheel but it is quite
impossible by means of bipedal gait.
The next most economical method would be
translation of the body through a sinusoidal
pathway of low amplitude in which the deflections
are gradual.”
8. • Inspired
• Elegant
• Persuasive
• Repeated in nearly all the major text books.
• Wrong!
9. Gard, S., & Childress, D. (1997). The effect of pelvic list on the vertical displacement of the
trunk during normal walking. Gait and Posture, 5:233-238.
Gard, S., & Childress, D. (1999). The influence of stance-phase knee flexion on the vertical
displacement of the trunk during normal walking. Arch Phys Med Rehabil, 80:26-32.
Kerrigan, D. C., Della Croce, U., Marciello, M., & Riley, P. O. (2000). A refined view of the
determinants of gait: significance of heel rise. Arch Phys Med Rehabil, 81(8), 1077-1080.
Gard, S., & Childress, D. (2001). What determines the vertical displacement of the body
during normal walking? Journal of Prosthetics and Orthotics, 13, 64-67.
Kerrigan, D., Riley, P., Lelas, J., & Della Croce, U. (2001). Quantification of pelvic rotation as
a determinant of gait. Arch Phys Med Rehabil, 82, 217-220.
Ortega, J. D., & Farley, C. T. (2005). Minimizing center of mass vertical movement increases
metabolic cost in walking. J Appl Physiol, 99(6), 2099-2107
Kuo, A. D. (2007). The six determinants of gait and the inverted pendulum analogy: A
dynamic walking perspective. Hum Mov Sci, 26(4), 617-656.
Gordon, K. E., Ferris, D. P., & Kuo, A. D. (2009). Metabolic and mechanical energy costs of
reducing vertical center of mass movement during gait. Arch Phys Med Rehabil, 90(1): 136-
144.
10. What went wrong?
• Cannot explain all of the features of
human gait on basis of one criteria
(smooth the trajectory of centre of mass).
• Inman and friends only ever thought about
the problem - they never tried to match
their conjectures to any data.
10
11. New approach
1. Identify the Requirements for walking
2. Start off with a simple model
3. Add in complexity that we understand
4. Test against our data
5. Keep adding complexity until we
understand the major features of human
walking
Work in progress
11
12. New approach
• Not quite as simple
• Not quite as elegant
• Not quite as wrong!
12
13. What are the requirements of
functional human walking?
13
14. Requisites of walking
Continuing ground reaction forces
that support the body
Periodic movement of each foot
from one position of support to the
next in the direction of progression
14
Inman V, Ralston H, & Todd F (1981). Human Locomotion.
?
16. Major motor functions
Maintenance of support
Maintenance of upright posture
Control of foot to achieve safe clearance and
a gentle heel or toe contact
Generation of energy to maintain the present
forward velocity or accelerate
Absorption of mechanical energy for shock
absorption and stability or decelerate
16
Winter D.A. (1991). Biomechanics of Human Gait
?
?
17. Pre-requisites of normal gait
Stability in stance
Clearance in swing
Pre-positioning in foot in terminal swing
Adequate step length
Energy conservation
17
Gage, J. (1991). Gait Analysis in Cerebral Palsy.
?
18. Requirements of functional walking
Energy conservation
Clearance in swing
Adequate step length
Support of bodyweight
Smooth transitions
18
Baker, R. (2009). Melbourne Gait Courses.
21. Loading
response
Initialcontact
Mid-stance Terminal stance Pre-swing Initial swing Mid-swing
Terminal
swing
0% 2% 10% 30% 50% 60% 73% 87% 100%
Initial contact
isn’t a phase
Loading occurs
throughout first
single support
Why not
pre-stance?
Single support and swing are
divided into a different number
of phases
Mid-stance isn’t
in the middle of
stance
Terminal stance
isn’t at the end
of stance
Pre-swing
emphasises
continuity of gait
cycle
Conventional terminology
Vertical component of ground reaction
22. Loading
response
Initialcontact
Mid-stance Terminal stance Pre-swing Initial swing Mid-swing
Terminal
swing
0% 2% 10% 30% 50% 60% 73% 87% 100%
New proposal
First
double
support
1DS
Second
Double
Support
2DS
Early
single
support
ESS
37%23%
Mid-
single
support
MSS
Late
single
support
LSS
Early
swing
ESw
Mid-
swing
MSw
Late
swing
LSw
24. Modelling
• Making simplifications and assumptions to be
able to understand something that is very
complex
• Stating explicitly what those assumptions are and
knowing the limitations of our conclusions.
• Making predictions on the basis of the model
• Testing those predictions against experimental
data
24
25. Assumption 1: We are working in two dimensionsAssumption 2: The human body can be modelled as a number
of rigid segments – Head/Arms/Trunk (HAT),
Femurs, Tibias and feet
Introducing our model: e-Verne
25
Assumption 3: The segments are linked by simple jointsAssumption 4: The movement about each joint can be
specified by a simple joint angle
26. 26
blog: www.wwRichard.net
Requirements of walking
Energy conservation
Clearance in swing
Appropriate step length
Support of bodyweight
Smooth transistions
Requirements of walking
Energy conservation
Clearance in swing
Appropriate step length
Support of bodyweight
Smooth transistions
27. Efficiency of walking
Walking for a kilometre at comfortable
speed (4kmh) uses up the energy in
two teaspoons of sugar.
A healthy child has to walk for over an
hour to work off the energy contained in
a can of coke.
27
34. Validation
10 50
1DS 2DS
60 1000
34
McGrath, M., Howard, D., & Baker, R. (2014). The strengths and weaknesses of inverted
pendulum models of human walking. Gait and Posture [elecrtronic publication]
35. Inverted pendulum
• Models horizontal components of velocity
well throughout cycle
• Models vertical component well through
single support (but not double support)
• Faster you walk the less good the inverted
pendulum model is.
35
40. Compass gait
a) No double support so stance and swing both
50% of gait cycle.
b) Can only see angles for one leg because in
walking is symmetrical.
c) Pelvis tilt fixed at 14° (because PSIS above ASIS)
d) Hip extends throughout stance
e) Hip flexes throughout swing
f) Femur movements are symetrical about vertical
but offset because of pelvis.
g) No knee movement
h) Ankles mirror hips exactly
i) Feet are horizontal as they scrape along floor.
f
c
d e
g
h
a
i
42. Clinical implications
• Walking is a dynamic activity requiring
preservation of kinetic energy from step to
step.
• It can’t be taught (re-taught) as a
sequence of static postures.
42
43. 43
blog: www.wwRichard.net
Requirements of walking
Energy conservation
Clearance in swing
Appropriate step length
Smooth transistions
Support of bodyweight
Requirements of walking
Energy conservation
Clearance in swing
Appropriate step length
Smooth transistions
Support of bodyweight
44. 44
blog: www.wwRichard.net
Requirements of walking
Energy conservation
Clearance in swing
Appropriate step length
Support of bodyweight
Smooth transistions
Requirements of walking
Energy conservation
Clearance in swing
Appropriate step length
Support of bodyweight
Smooth transistions
45. What is the minimum knee flexion
required for clearance with a
plantigrade ankle?
Interactive poll
45
50. Clearance
j) Minimum toe clearance occurs about half way
through swing
k) Knee flexes to at least 50° by mid swing
l) Hip must be flexed early in swing
m) Ankle must be at least neutral in mid-swiing
n) Foot will follow knee (moderated by hip flexion)
j
k
l
m
n
51. Clinical implications
• If you are going to use kinee flexion to
clear the leg you need a lot of it
• Small amounts make things worse
• Modifications of hip and ankle movement
are required for knee flexion to be effective
51
52. 52
blog: www.wwRichard.net
Requirements of walking
Energy conservation
Clearance in swing
Appropriate step length
Smooth transistions
Support of bodyweight
Requirements of walking
Energy conservation
Clearance in swing
Appropriate step length
Smooth transistions
Support of bodyweight
53. What are the four most effective
joints for increasing step length
Interactive poll
53
59. Step length
o) Step length primarily determined by difference
in hip flexion-extension between opposite foot
contact and foot contact.
p) Require good leading knee extension at foot
contact.
q) Require some trailing knee flexion before
opposite foot off.
r) Requires a little heel rise to facilitate knee
flexion
p
o
q
r
60. Clinical implications
• Step length is fundamentally determined
by the range of hip movement
• Extension of the leading knee and flexion
of the trailing hip are also important
• Ankle movement and heel rise play a
minimal role in determining step length
60
61. 61
blog: www.wwRichard.net
Requirements of walking
Energy conservation
Clearance in swing
Appropriate step length
Smooth transistions
Support of bodyweight
Requirements of walking
Energy conservation
Clearance in swing
Appropriate step length
Smooth transistions
Support of bodyweight
62. 62
Upward velocity
• Upward velocity reduces throughout
• There is a downward acceleration
• The ground reaction must be less than gravity
69. Support of Bodyweight
The inverted pendulum motion requires a double
support phase.
r. Stance must thus be longer than swing.
s. Opposite foot contact and opposite foot off
become meaningful.
70. 70
blog: www.wwRichard.net
Requirements of walking
Energy conservation
Clearance in swing
Appropriate step length
Smooth transistions
Support of bodyweight
Requirements of walking
Energy conservation
Clearance in swing
Appropriate step length
Smooth transistions
Support of bodyweight
71. Two transitions
1. Stance to swing at for off
2. Swing to stance at foot contact
“It’s a lot easier to fall off a log than onto one”
Richard Baker – August 2009
The swing to stance transition is by far the
more difficult
71
73. 73
Winter DA. Foot trajectory in human gait: a precise and multifactorial motor control task.
Phys Ther. 1992;72(1):45-53;
“The trajectory velocity of the heel immediately prior to HC is
virtually zero vertically …
… and low in the horizontal direction;
such findings raise the question as to why many researchers
refer to this initial contact as ‘heel-strike’”
74. David Winter
“Primary tasks of walking:
3) control of the foot trajectory to achieve
safe ground clearance and a gentle heel
or toe landing."
74
Winter DA. Biomechanics and motor control of human movement.
Third Edition, John Wiley and Sons, Hoboken, New Jersey, 2004
75. Smooth transition - horizontal
75
Heel speed is less than 5% of
maximum at foot contact
(Winter exaggerated this by
measuring ankle speed)
Swing Stance
Pongmala et al. Is foot contact a collision? GCMAS 2015
Slow, normal and fast
speeds
78. Horizontal – late swing
Achieved through swing limb mechanisms:
1. Knee flexion before foot contact
2. Plantarflexion before foot contact
You don’t read this in the text books!
78
80. Horizontal – early stance
Late swing motion is continued
1. Knee continues to flex
2. Ankle continues to plantarflex
Knee flexion in late swing and early stance
serves to avoid “shock” not to absorb it.
80
81. 81
Smooth transitions -
horizontal
r) Knee flexes before initial contact and continues
into early stance.
s) Ankle has to be approximately neutral and
plantarflexing prior to foot contact and this
continues in early stance.
t) Foot angle is modified by changes in knee and
ankle in late swing and comes down to
horizontal in early stance.
rr
s
s
t
t
82. Smooth transitions – Centre of mass
82
Centre of Mass moving
at maximum speed
1.5 m/s.
Horizontal velocity
87. Trailing limb must get longer during late stance
and 2nd double support
1. Plantarflexion
2. Controlled knee flexion (reduces leg length)
87
Smooth transitions – Centre of mass
90. Leading limb must get shorter during 1st
double support
1. Stance phase knee flexion
90
Smooth transitions – Centre of mass
91. Smooth transitions -
vertical
U. Limit dorsiflexion in late single support
V. Plantarflexion through double support (to
maintain length of limb as knee flexes)
91
u
v
92. Clinical implications
• Knee flexion in late swing is essential to
avoid “shock” at contact
• Prosthetic limbs have no mechanism for
this and hence heavy damping of impact is
required.
• Excellent motor control is required to avoid
shock. Toe walking may be a much
simpler mechanism if this is absent.
92
93. Which bump does what?
93
Maxvelocity
Zerodvelocity
The second peak of the ground reaction slows the body down in a
vertical direction – “push-off” is an extremely mis-leading term
Dedeleration
94. Which bump does what?
94
Maxvelocity
Zerodvelocity
The first peak of the ground reaction is when the body is
accelerated upwards.
Acceleration
96. Which bump does what?
• In the horizontal direction the opposite
happens:
– Early stance is a deceleration phase
– Late stance is an acceleration phase
96
98. Original aim
1. Identify the Requirements for walking
2. Start off with a simple model
3. Add in complexity that we understand
4. Test against our data
5. Keep adding complexity until we
understand the major features of human
walking
98
101. 101
Requirements of walking
Energy conservation
Clearance in swing
Appropriate step length
Support of bodyweight
Smooth transistions
Can we apply these principles to
understand walking with pathology?
Translation of the body in a straight line with the least expenditure of energy may be achieved mechanically by means of a wheel but it is quite impossible by means of bipedal gait
The next most economical method would be translation of the body through a sinusoidal pathway of low amplitude in which the deflections are gradual.
Import Excel graph as Enhanced metafile, group and convert to Microsoft graphic object.
(Make sure that the horizontal axis is not leading to cropping of series data).
Height 540
Width 620
Left 60
Height 540
Width 620
Left 60
These are the graphs of how the horizontal (top) and vertical (bottom) components of the speed of the centre of mass change as the inverted pendulum operates. The CM has to be travelling reasonably quickly in a forwards direction at the (1) start of the movement because it is going to slow down as gets higher and kinetic energy is converted to potential energy. It will be going slowest in the (2) middle of the movement when the mass is highest after which it picks up (3) speed again as the potential energy is converted back to kinetic.
We can also think about what is happening in a vertical direction. At the start of the (4) movement the centre of mass is moving upwards. As it moves around the circular arc that’s dictated by the hip joint rotating about the ankle then two things happen. In the first half of the movement the body is losing kinetic energy so the vertical as well as the horizontal component of velocity must reduce, but also the direction of movement flattens out so a smaller part of that movement is in the upwards direction. The two effects reinforce each other so that the vertical velocity reduces through the first half of the movement. It is actually zero as the pendulum (5) passes over its highest point because, just for a split second, the centre of mass is actually travelling exactly horizontally. We can put this a different way and say that the centre of mass is decelerating in a vertical direction over the first half of the movement.
After passing the mid-point then the same two factors act in the opposite sens (6) e. The body gains kinetic energy so starts moving faster and as it moves round the arc a greater proportion of its velocity is directed downwards. Both factors combine and the vertical component of velocity becomes increasingly negative (it is accelerating downwards).
(1) Now we said we were going to talk about forces and indeed we already are because Newton’s laws tell us that if accelerations are occurring then forces must be acting. (2) Let’s think about what those forces must be. We’ll look at the vertical component of the ground reaction on this graph with (3) bodyweight as our reference.
Early on (4) the body is travelling upwards (positive vertical component) but getting slower – it’s decelerating. The overall force on the body must be acting downwards. That is the (5) vertical ground reaction acting upwards must thus be less than bodyweight acting downwards. You can see that the slope of the velocity graph is getting a little more gradual with time so the ground reaction must be getting closer to bodyweight with time.
In the (6) middle of the movement the body goes over the top of the arc and from moving upwards to moving downwards so it is still accelerating downwards. (7) The ground reaction must still be less than bodyweight. The slope is at its gentlest at this point though so the reaction will be closest to bodyweight at this point.
In the (8) later part of the movement the centre of mass is accelerating downwards so again the (9) ground reaction must be less than bodyweight and as the slope gets steeper it must the ground reaction must be getting smaller.
You can see that the ground reaction under an inverted pendulum is always less than gravity. I haven’t been able to trace where but I think Jacquelin Perry once referred to walking as a process of continuous falling – I don’t know if she knew quite how biomechanically accurate.
Most of us will know that the ground reaction doesn’t actually look like that under the inverted pendulum, it’s got two (1) characteristic bumps. These certainly increase the ground reaction quite considerably above that generated by the inverted pendulum (we’ll discuss the mechanisms for this in a later screen cast) but look closely at the graph in relation to bodyweight. The ground reaction is only (2) above bodyweight is only above bodyweight for a fairly short period and by not very much.
If we contrast this with the amount of time that the ground reaction is less than bodyweight and the size of the deficit you’ll see that on average the ground reaction generated is considerably less than bodyweight. We are still not supporting bodyweight through stance.
Here is the data plotted for left and right limbs you’ll see that the average vertical component of the ground reaction (the horizontal green line) is a little over 80% of bodyweight.
If we allow a short period of double support however …
… you’ll see that the ground reaction under the left and right limbs combines and that the greatest total ground reaction acting on the body is actually just a little before the middle of double support when two relatively modest individual ground reactions combine. The result of this is that the average value of the ground reaction is thus increased. We thus reach the conclusion that if we are going to rely on the inverted pendulum as the essential mechanism for moving the body forwards then we will have to have a period of double support to ensure that bodyweight is supported.
If the ground reaction was purely that under an inverted pendulum you can work out that the double support period would have to be about 15% of the gait cycle. With the modified form of the ground reaction we find in normal walking it is about 10%.
The only experimental work I am aware of that has sought to quantify the extent to which foot contact can be considered a collision is David Winter’s work from over 20 years ago. He measured a “virtually zero” [1] component of velocity in the vertical direction and a low velocity [2] in the horizontal direction which led him to conclude that use of the term “heel-stike” was inappropriate.
In a later work he described five major motor tasks of which the third was “Control of the foot trajectory to achieve safe ground clearance and a gentle heel or toe landing."
He also framed five primary tasks of walking of which the third included controlling the foot trajectory to give a gentle heel or toe landing.
As we commented in an earlier screen cast the centre of mass has to be travelling at maximum speed at the same time in order that the kinetic energy from one gait cycle can be carried over to the next gait cycle whilst the potential energy is at a minimum. It’s thus travelling at around 1.5m/s at this point.