Note: Descriptions are shown in the official language in which they were submitted.
CA 02406622 2002-10-23
WO 01/85402 PCT/CA01/00678
MOTION PLATFORM AND METHOD OF USE
Inventors: Keith G. Chrystall, Simon P. Monckton, Mark Oleson
Assignee: Alberta Research Council Inc.
FIELD OF THE INVENTION
The present invention relates to a motion platform and in particular, to a
method
and apparatus incorporating a motion platform to test shoes.
BACKGROUND OF THE INVENTION
Motion platforms are devices that can move an object through an arbitrary
trajectory in 3 dimensional space within a given working volume while
maintaining precise
control over the position, velocity and acceleration of the object. One type
of mechanism
that is suitable for creating a general-purpose motion platform is a Stewart
platform or
hexapod. Stewart platforms are well-known and include a parallel linkage that
permits
movement with 6 degrees of freedom: independent and simultaneous translation
and
rotation about and along each of the three primary Cartesian axes X, Y and Z.
A Stewart
platform includes six "legs" each of which is extendible to translate and
rotate a platfonm
about all three Cartesian axes. In general, Stewart platforms are known for
their law mass,
high mechanical stiffness and large payload capacities.
Typically, Stewart platforms have extendible "legs" which are hydraulic,
pneumatic, lead screw electric or ball screw electric actuators and are used
in automated
assembly lines. Examples of automatic assembly machines using a Stewart
platform are
disclosed in U.S. Patent Nos. 6,041,500 and 5,987,726. These prior art Stewart
platforms
generate tremendous thrust, particularly hydraulically actuated platforms, at
the expense of
velocity and are limited to accuracy in the range of one to two millimetres.
In many applications, it is desirable to generate platform velocities of
greater than
2.5 metres per second and up to 3 m/s or more, with accuracy in the sub-
millimetre range
CA 02406622 2002-10-23
WO 01/85402 PCT/CA01/00678
and maintain thrust in the range of 1000 pounds. Further it would be desirable
if such a
platform could also deliver vertical acceleration in excess of 3 g and
rotation rates of up to
1000 degrees per second. One such application is footwear testing.
It is desirable to determine the effect of shoe design on foot motion and
impact
control, particularly for athletic footwear. Athletic footwear manufacturers
typically
evaluate human subjects wearing test shoes during prescribed movements. These
human
tests are time-consuming, unreliable and fail to reveal internal motion or
forces within the
ankle or leg, apart from subjective observations by the test subject, which
are inherently
unreliable.
Robotic footwear testers have been directed at simulating wallcing or running
motions. One example is the tester disclosed in U.S. Patent No. 4,130,007. The
difficulty
with. such robotic testers is that they are mechanically complex yet only very
simple
wallcing or running motions may be simulated and they cannot accurately
simulate
complex motions. As a result, only very limited test information is obtained.
Therefore, there is a need in the art for methods and apparatuses to
robotically test
footwear which may incorporate a novel motion platform.
SUMMARY OF THE INVENTION
The present invention is directed to a novel motion platform falling within
the
generic category of Stewart platforms or hexapods. In particular, the motion
platform of
the present invention is incorporated into a novel apparatus and method for
testing
footwear.
Therefore, in one aspect, the invention comprises a motion platform comprising
a
base, a platform, a motion controller and at least 6 linear actuating
mechanisms disposed
between and connected to the base and the platform, wherein each actuating
mechanism
comprises an electromagnetic linear actuator comprising an electric thrust
block and a
magnetic thrust rod and each actuating mechanism is controlled by the motion
controller.
-2
CA 02406622 2002-10-23
WO 01/85402 PCT/CA01/00678
In another aspect, the invention may comprise a shoe testing apparatus
comprising:
(a) a fixed artificial foot including a force sensor, for wearing the shoe;
(b) a motion platform capable of moving and being cantrolled with 6 degrees of
freedom;
(c) means for actuating the motion platform and striking the shoe with the
platform in a manner simulating the running or walking gait of a shoe
wearer;
(d) means for recording the forces sensed and transmitted by the force sensor.
In another aspect, the invention may comprise a method of testing a shoe,
comprising the steps of:
(a) fitting an artificial foot within the shoe, providing a force sensor
attached to
the artificial foot and fixing the foot and shoe in position;
(b) providing a motion platform capable of moving and being controlled in 6
degrees of freedom;
(c) providing a foot trajectory and transforming the foot trajectory to a
platform
traj ectory;
(d) actuating the motion platform with the platform trajectory to strike the
shoe;
and
(e) recording the resulting forces with the force sensor.
-3
CA 02406622 2002-10-23
WO 01/85402 PCT/CA01/00678
BRIEF DESCRIPTION OF THE DRAWINGS
The invention will now be described by way of an exemplary embodiment with
reference to the accompanying simplified, diagrammatic, not-to-scale drawings.
In the
drawings:
Figure 1 is a perspective view of one embodiment of a motion platform of the
present invention;
Figure 1A is a perspective view of one embodiment illustrating the crash ring;
Figure 2 is a view of one linear actuator of one embodiment;
Figure 3 is an exploded view of a bearing and bearing housing of one
embodiment
of a linear actuator;
Figure 4 is an exploded view of a rod end clamp;
Figure 5 is a top plan view of the base plate. Figure SA is a side view of the
base
plate;
Figure 6 is a bottom view of the strike plate. Figure 6A is a side view of the
strike
plate;
Figure 7 is a top view of the strike~plate superimposed over the base plate.
Figure
7A is a schematic side view of the strike plate and base plate.
Figure 8 is a schematic flowchart of one embodiment of a method of the present
invention.
-4
CA 02406622 2002-10-23
WO 01/85402 PCT/CA01/00678
Figure 9 is a schematic representation of one embodiment of a footwear test
apparatus of the present invention.
Figure 10 is a screen shot of exemplary output graphs from a footwear test of
the
present invention.
DETAILED DESCRIPTION OF THE INVENTION
The present invention provides for a motion platform and a method of using a
motion platform to test footwear. When describing the present invention, the
following
terms have the following meanings, unless indicated otherwise. All terms not
defined
herein have their common art-recognized meanings.
The term "Stewart platform" or "hexapod" refers to an octahedral platform
including stiff extendible legs actuating a platform in alI six degrees of
freedom: translation
and rotation about three Cartesian axes. Each leg is attached to a base and
the platform by
flexible joints.
The term "comprises" or "compzising" or "includes" or "including" shall mean
"includes without limitation" or "including without lunitation" as the case
may be.
The apparatus ( 1 ) shown in the Figures comprises a base plate (7) which may
be rigidly
fastened to a massive object, such as the ground or floor of a building. It
can be fastened to
the walls or ceiling of a structure providing that these surfaces are strong
enough to provide
complete rigidity through the operating range of the motion platform. Mounting
adapters (5)
and actuator joints (6) connect linear actuators (8) to the base plate (7).
The mounting
arrangement is identical for each of the six linear actuators (8). The
mounting adapter (5) is
designed to maintain the orientation and location of the rotational centre of
joint (6) in a
precise pose with respect to the geometric centre of the base plate (7). In
one embodiment, the
design of mounting adapter (5) and its placement position on base plate (7) is
governed by a
set of relationships and methodologies that relate the machine configuration
to the desired
woxking volume and range of movement of the motion platform (1). These
relationships and
-5
CA 02406622 2002-10-23
WO 01/85402 PCT/CA01/00678
methodologies are detailed below.
Joint (6) must provide at least two independent axes of rotation and the axes
of rotation
must intersect such that there is a centre of rotation that does not move with
respect to the base
plate (7). In one embodiment, commercially available universal joints (Item
Number UJ-
NB435-1004 from Belden Inc. of Broadview, Illinois) were used for joint (6).
Other
commercially available or custom designed joints could be used providing they
satisfy the
principle requirement regarding a non-moving center of rotation and have
strength, wear and
friction characteristics that are compatible with the intended application.
Joints (4) and mounting adapters (3) connect the linear actuators (8) to the
strike plate
(2). The requirements for j oints (4) and mounting adapters (3) are the same
as the requirements
for the joints (6) and the mounting adapters (5) on the base plate (7). In one
embodiment,
joints (4) and mounting adapters (3) on the strike plate (2) are identical to
the base plate joints
(6) and mounting adapters. However, the strike plate mounting adapters (3) and
joints (4)
need not be identical to their counterparts on the base plate (7) so long as
they satisfy the
configuration requirements for the motion platform (1) and are suitable for
the intended
application.
The placement position for mounting adapter (3) on strike plate (2) are
governed by a
set of relationships and methodologies that relate machine configuration to
the desired
working volume and range of movement of the motion platform (1). These are the
same
relationships and methodologies that determine the design of mounting adapters
(3, 5) and
determine the placement location of mounting adapter (3) on base plate (7) and
are detailed
below.
Strike plate (2) provides a surface on which the object to be moved can be
mounted.
Examples of such objects are cameras, communication antennas, people and
articles to be
tested through a motion sequence. In one embodiment, the strike plate (2)
serves as a
mounting surface for a carpet material, and is moved to strike a stationary
shoe in a manner
that replicates the kinematics of human running. For this application and
others, it is
-6
CA 02406622 2002-10-23
WO 01/85402 PCT/CA01/00678
preferable that the strike plate (2), mounting adapter (3) and joint (4) are
all manufactured
from aluminum to minimize the 'flying' weight of motion platform (1). These
components
can be made from materials other than aluminum depending on the needs of the
intended
application.
The motion platform may include an additional component referred to as a
"crash ring"
(30). The crash ring (30) is a passive component that encircles the motion
platform at a
distance that does not interfere with normal movement. The crash ring (30)
acts as a physical
stop for the motion platform during situations where the device topples over
to the side of its
work envelope. This situation normally occurs when the motion platform loses
power when it
is displaced laterally from its center position. The crash ring prevents
physical damage to the
motion platform during these incidents and allows the motion platform to be
quickly returned
to its centered home position.
The operating characteristics for the linear actuator (8) are determined by a
set of
relationships and methodologies that relate machine performance to the
position, velocity and
acceleration characteristics of the trajectories to be followed by the motion
platform (1). In
one embodiment, the linear actuator (8) may be a linear electric jack. Other
types of custom
designed or commercially available electric, hydraulic or pneumatic actuators
could be used
providing that they permit movement only along one axis and that they satisfy
the motion
requirements for the intended application.
Referring to Figure 2, the linear motor assembly (8) is composed of a number
of key
sub-components: a linear electric motor block (9) and magnetic thrust rod
(19); bearing
assemblies (10); side rails (29, 30) and bottom plate (26); rod end clamps
(31); encoder (17),
encoder mounts (15 16); and encoder connecting rod (20).
The linear electric j ack relies on a linear electric motor, composed of a
powered linear
thrust block and a thrust rod (19) which is a stainless steel tube packed with
magnets. In one
embodiment of the linear electric jack, the LD3808 linear electric motor block
was used in
conjunction with a 38mm thrust tube, which are both commercially available
from Linear
_7
CA 02406622 2002-10-23
WO 01/85402 PCT/CA01/00678
Drives of Basildon, Essex, U.K.
The motor housing may be comprised of two supporting aluminum rails (29, 30)
and
an end plate (26). Tlus open structure pernzits internal cooling within the
motors and, with
appropriate material selection, a strong load path that supports anticipated
loads, the motor,
bearing housings (10), linear measurement system (17), and any other external
mounting
requirements. The end plate (26) completes the structure and provides an
additional coupling
surface for externally mounted components. In one embodiment, gauge aluminum
plate was
used to make the rails (29, 30) and end plate (26). If appropriately
dimensioned for strength
and cooling, other materials including plastic, steel or composites (such as
fiberglass) could
meet the requirements of this linear electric j ack.
The magnetic thrust rod (19) is driven linearly through the hollow motor block
(9)
when appropriately commutated current is applied to the motor. The rod (19) is
held clear of
the motor block's inner wall by two bearing assemblies (10) on either side of
the thrust block.
Referring to Figure 3, the bearings (13) are held rigidly by bearing housings
(11,12) and a
bearing retaining ring (14). For linear electric motors, these bearings should
preferably be dry
(unlubricated) and made of a material with properties such as low back
electromotive force
(EMF) and low friction. Such materials may include carbon or plastic. The
bearing housings
should also be manufactured using material with low back EMF properties such
as plastic or
brass. Both the bearings (13) and the bearing housings (11, 12) may be split
to facilitate ira situ
removal and/or replacement of bearing components without disassembly of the
linear electric
jack. In one embodiment of such split bearing assemblies, the bearings (13)
were composed of
a Carbon/Resin composite, a dry running low friction material, and custom
manufactured by
Advanced Carbon Products of Hayward, California, USA. In this same embodiment,
the
bearing housings were made from ErtaliteTM, a light, strong, plastic with no
back EMF
characteristics manufactured by DSM Engineering Plastic Products of Reading,
Pennsylvania,
USA.
Rod end clamps (20, 31) permit the attachment of payloads to the thrust rod
(19) and,
for the interior rod end clamp (20), prevents the thrust rod (19) from exiting
the motor
_8
CA 02406622 2002-10-23
WO 01/85402 PCT/CA01/00678
assembly. Referring to Figure 4, the rod end clamps (31) are designed to
provide solid
attachment points to the rod without machining the rod surface or end
components and to
provide a solid impact surface during overstrokes. The clamps may comprise a
split clamp ring
(21), a clamp collar (22), and a clamp compression plate (34). The thrust rod
(19) is inseued
into the clamp collar (22) and clamp ring (21) such that the wedge profile of
the ring fits
snugly into the collar. The compression plate (34) is then fastened to the
clamp collar using a
number of appropriately sized industrial fasteners. As these fasteners are
tightened, the clamp
collar (22) and clamp ring (21) are pulled against the compression plate (34).
The wedge
profile of the clamp ring (21) and collar (22) drives the clamp ring (21)
against the rod (19)
surface. As the clamp ring (21) is compressed, the compression plate (34) is
pulled against the
rod (19) end. In this embodiment, the compression plate (34) provides both a
firm payload
mount and a reliable geometric reference relative to the rod end (19) which is
important for
precision motion. The angle of the wedgeA is related to clamping force, Nb ,
and the
coefficient of static friction, ~.~ , through an explicit relationship
expressed below:
N - Fb(~ose -wa Sme> (1)
sine +~a cos6
To provide closed loop control of the motor, the control system must receive
the motor
position as feedback. There are many encoders (15) which are commercially
available and
suitable for use with the present invention. Any linear measurement system of
appropriate
resolution could be used for this task, however many suffer from problems
common to linear
measurement systems requiring moving power and signal connections. These
latter systems
should preferably be avoided. In one embodiment, a magneto-strictive encoder
was used.
These devices are typically composed of a passively magnetic encoder carriage
(18) and an
encoder body (17), containing magnetically sensitive material through which
the time-of
transit of an ultrasonic pulse is measured and converted into displacement by
means of internal
electronics. In one embodiment, the LP-38 ISI magneto-strictive encoder from
TRelectronic
GmbH of Trossingen, Germany was selected to provide A quad B incremental
output to the
system's control electronics. The control electronics in one embodiment of the
present
invention may be an IDC B8001 servodrive from Industrial Devices Corporation
of Petaluma,
-9
CA 02406622 2002-10-23
WO 01/85402 PCT/CA01/00678
California, and a Delta Tau PMAC1 by Delta Tau Data Systems of Northridge,
California,
USA.
The encoder carriage (18) limited to strictly linear motion can be connected
to the
interior rod end (19) through an encoder connector assembly (20). The encoder
connector
assembly is composed of a modified rod end compression plate (23) with a short
shaft, a
bearing (24), and a connector rod (25). The connector rod (25), bearing (24)
and compression
plate shaft (23) are j oined by press fitting, for example, into a single
assembly that permits free
rotation of the connector rod about the shaft of the compression plate (23).
Therefore, the
encoder connector assembly permits the thrust rod (19) to rotate freely about
the rod axis
without applying detrimental forces to the encoder while communicating rod
linear motion
rigidly to the linear encoder (17). Encoder mounting plates (15,16) provide an
offset distance
for the linear encoder, typically dependent on the encoder type and EM
sensitivity properties
of the encoder. Other linear measurement systems such as LVDTs, laser
interferometers, and
glass scale encoders might also be applied to this system and, depending on
the placement and
measurement principle, might or might not require specific rod end couplers.
Unlike some rotary motors, linear motors are capable of exceeding mechanical
limits
and damaging components in the process, an event called "overstroke". To
prevent this
possibility, the actuator (8) may use positive and negative limit switches
(28) that disable the
motor when triggered by the entry of the rod into a dangerous region. In one
embodiment,
Banner D 12 optical switches by Banner Engineering Corporation of Minneapolis,
Minnesota,
monitor and enforce the linear actuators interior geometric limits and disable
the amplifiers on
overstroke. Such switches (28) may be further used to implement homing
triggers for precise
calibration of the motor position.
The damage from an overstroke event may further reduced through the selection
of
appropriate end plate bumper (27) and rod end ring bumpers. In one example,
the rod end ring
bumper may be manufactured of neoprene while the lower bumper may be made of
40 duro
gum rubber . The rod end ring bumper can be fixed either to the bearing
housing (10) or rod
end clamp collar (22) as convenient.
-10
CA 02406622 2002-10-23
WO 01/85402 PCT/CA01/00678
The relationships and methodologies that relate machine configuration to the
desired
working volume and range of motion of the motion platform (1) are now
described. The
location of the center of rotation of j oint (6) on base plate (7) is governed
by three independent
parameters that are measured from the geometric center of base plate (7) as
illustrated in
Figure 5:
PMP X - Primary Mounting Point X value
PMP Y - Primary Mounting Point Y value
PMP Z - Primary Mounting Point Z value
The location of the center of rotation of j oint (4) on strike plate (2) is
governed by three
independent scaling factors that are applied respectively to PMP X, PMP Y and
PMP Z as
illustrated in Figure 6:
X DFF - X Dimension Form Factor
Y DFF - Y Dimension Form Factor
Z DFF - Z Dimension Form Factor
The location of the center of rotation of joint (4) on strike plate (2) is
measured from
the geometric center of strike plate (2). The geometric center of strike plate
(2) is itself
centered on the geometric center of base plate (7) with an offset in the Z
(vertical) direction.
The length of linear actuator (8) determines the linear distance between the
center of
rotation of joint (6) on base plate (7) and the center of rotation of joint
(4) on strike plate (2).
Once the value for the length of the linear actuator is established and values
are set for
PMP X, PMP-Y, PMP Z, X DFF, Y DFF and Z DFF the configuration of motion
platform
(1) is calculated and its corresponding range of motion and working volume is
established.
The azimuth (0) and altitude (~) angles (see figure 7) of mounting adapters
(3, 5) are
calculated using the following relationships:
-11
CA 02406622 2002-10-23
WO 01/85402 PCT/CA01/00678
8 = ATAN2~((PMP Xl2-~3*PMP Yl2)+X DFF*PMP ~, (Y DFF*PMP Y
(~3 *PMP Xl2+PMP Yl2))J
(2)
~= ACOS~LISJ
(3)
where:
S = the neutral or homed length of linear actuator (8); and
L=(S~ - ((PMP Xl2-~I3*PMP Yl2)+X DFF*PMP ~2 -(Y DFF*PMP Y
(~I3 *PMP-Xl2+PMP Yl2))~)1 /2
This design insures that linear actuator (8) and its corresponding joints are
held inline
when the linear actuator (8) is in its "homed" or neutral position (as defined
by the length
value for the Linear Actuator used in the above calculations). By maintaining
this 'inline"
configuration the maximum working volume for motion platform ( 1 ) is realized
with a given
joints (4, 6).
Having a desired trajectory for strike plate (2) allows the calculation of the
corresponding movements that each of the six linear actuators (8) must make in
order to
achieve this motion. The desired traj ectory is represented mathematically by
a matrix of 4 x 4
transforms that represent the position and orientation (i.e. the 'pose') of
strike plate (7) through
the duration of the trajectory. Each element of the trajectory matrix is a 4 x
4 transform that
represents the pose of strike plate (7) at a moment in time. The first element
of the trajectory
matrix is the pose of strike plate (7) at the begim~ing of the trajectory.
Each subsequent
element of the traj ectory matrix represents the pose of strike plate (7) at a
specific interval of
time after the beginnW g of the traj ectory. The last element of the traj
ectory matrix represents
the pose of strike plate (7) at the end of the trajectory. The positions of
the centers of rotation
of each of the joints (4) can be calculated using the following formula:
-12
CA 02406622 2002-10-23
WO 01/85402 PCT/CA01/00678
pr=t = T =t ' .pr=o
Where: p = a 6x4 array representing the positions of the centers of rotation
of joints (4) in
Cartesian coordinates
T = the traj ectory matrix
t = time
i = a time interval with the trajectory matrix
't=0' represents the beginning of the trajectory
The length of each of the linear actutators (8) at each time interval along
trajectory can
then be calculated by subtracting the position of the center of rotation of
joint (6) from the
position of the center of rotation of j oint (4) attached to the linear
actuators (8). The formula
for this calculation is:
'natr=r = natr=t -Bn natr=i -Bn
Where: lye at t=i - is the length of 'n'th linear actuator (8), at the 'i'th
time interval along the
traj ectory and 'n' ranges from 1 to 6 for each of the six linear actuators
(8)
P- is the position of the center of rotation of joint (4) in Cartesian
coordinates on the
'n'th linear actuator (8) at the 'i'th time interval
B - is the position of the center of rotation of joint (6) in Cartesian
coordinates on the
'n'th linear actuator (8) at the 'i'th time interval
Having established the movements for each of the linear actuators (8) it can
then be
determined if the joints (4, 6) can accommodate this motion. A mathematical
model of the
motion platform is used to determine the corresponding motions of joints (4,
6). In one
embodiment of the current invention a software package called "Working Model"
(The
MacNeal-Schwendler Corp. of San Mateo, California) was used to build the
mathematical
model of motion platform (1) and carry out the analysis. Other means of
numerical analysis
could be used to gain the same result. Depending on the outcome of the
analysis on the
movements of joints (4, 6) the parameters PMP X, PMP Y, PMP Z, X DFF, Y DFF
and
Z DFF, as well as the length of linear actuator (8), can be adjusted until an
appropriate
configuration is found to suit the desired trajectory(s).
-13
CA 02406622 2002-10-23
WO 01/85402 PCT/CA01/00678
The configuration of motion platform (1) may be readily adapted by those
skilled in
the art to a wide variety of other applications, such as mufti-degree of
freedom material
testing. In one embodiment of the invention, a specific example is the testing
of shoes to
determine the effect shoe design has on motion of the human ankle. In this
application, a
test shoe is held stationary by appropriate fixturing while the motion
platform becomes the
'ground' . A movement path, or traj ectory, is determined for the motion
platform that
creates the same relative motion between foot and ground as is experienced in
normal
human running or walking movement. Instrumentation mounted either to the
ground or to
the fixturing holding the test shoe measures the response of the shoe specimen
to the
application of the trajectory. These measurements can then be interpreted to
reveal the
desired information regarding the projected impact of the specimen on human
movement.
A schematic flowchart of one method of the present invention is shown in
Figure 8.
During a typical athletic movement, the motion of a human foot with respect to
a fixed
point on the ground and expressed over time is known as a foot trajectory.
This trajectory
can be either measured from human subj ects or generated through computer
simulation. A
foot trajectory may be transformed into a trajectory of the ground relative to
the foot using
computer software following the calculations outlined herein. Additional
"inverse
kinematics" software can transform this trajectory into a set of six linear
actuator
trajectories for a motion platform. These trajectories can then be formatted
into instructions
for an appropriate off the-shelf or customized motion platform controller.
Once executed,
these instructions can move the strike plate into the footwear 'specimen. A
six degree of
freedom force sensor in the "ankle" (or other location) of the specimen can
then be used for
force data capture. Together this process enables the discovery of forces
experienced by the
foot given a specific foot trajectory and shoe specimen.
Because the foot translates and rotates about all three Cartesian axes, a foot
trajectory is a six degree of freedom trajectory. To describe a "pose" at a
particular time in
this trajectory requires a position vector and some expression of orientation
(e.g.
orientation matrices, quaternions, eider angles, etc.). The trajectory is,
therefore, an ordered
-14
CA 02406622 2002-10-23
WO 01/85402 PCT/CA01/00678
set of such "poses" over time. The pose of a point on the foot with respect to
some
Cartesian coordinate system fixed to the ground can be represented
mathematically by a
4X4 matrix that incorporates both position and orientation information. The
4X4 matrix is
refeiTed to as a 'transform' and is symbolically represented as:
Gcs T
FCS
where: T- is a 4X4 numeric matrix
GGS- is the reference Ground Coordinate System
FCS - is a point attached rigidly to the foot and the Foot Coordinate System
A foot trajectory over time can therefore be represented as:
GCS
Fcs Tt
where - i represents the i~' time interval (0,1,2, . . ..)
The foot trajectory can be measured from human subjects using machine vision
techniques (e.g. Motion Analysis Corporation of 3617 Westwind Blvd., Santa
Rosa,
California 95403). The Cartesian position of reflective targets, applied to a
human subject,
are measured using software that processes images from a system of high speed
digital
cameras. The positions of these targets are then used to establish the
kinematics of motion
for the subject. Application of targets to known locations on the human foot
therefore
permits the measurement of the foot trajectory.
Foot motion can be simulated through the use of dynamic analysis software such
as
LMS DADS from LMS INTERNATIQNAL, Researchpark Z1 ,Interleuvenlaan 68 ,3001
Leuven, Belgium. Given the dimensions, mass and inertia of the human skeleton
and
muscle tissue, methods of human gait generation can be explored and an
appropriate foot
trajectory produced.
The foot trajectory may then be used to calculate a motion platform
trajectory. The
-15
CA 02406622 2002-10-23
WO 01/85402 PCT/CA01/00678
motion platform trajectory matrix, represented as T in the prior discussion of
the motion
platform design, can be calculated by the following formula:
-BCS (GCS 1 GCS
Tt- FcsT' CFCSTa ~ ' rcsT
where: BCS- is the Base Coordinate System for the motion platform;
PCS - is the strike Plate Coordinate System, which is rigidly attached to the
strike
plate; and
z - represents the i~' tune interval (0,1,2, . . ..)
The transform FcsT describes the pose of the Foot Coordinate System (FCS)
within
the Base Coordinate System framework. Since the pose of the FCS within the BCS
is
stationary for any given movement of the motion platform this transform can be
deduced
from the geometry of the motion platform, the fixturing holding the test shoe
and the
placement of these two items with respect to each other.
The Ground Coordinate System (GCS) framework is attached rigidly to the strike
Plate Coordinate System (PCS) since the strike plate is acting as the 'ground'
in this
application. The transform PcsT describes the pose of the PCS within the GCS
framework. This transform is not time dependent and can be deduced from the
geometry
of the strike plate and parameters established by the user.
The transform FcsT that represents the pose of the foot in the GCS must be
inverted
for this calculation using ordinary methods for inverting matrices. A 4X4
matrix that is
structured correctly to represent pose data is normally invertable.
As shown in Figure 9, one embodiment of the footwear test apparatus is
configured
to discover ankle force data from a foot traj ectory within the motion
platform system. The
apparatus may comprise a computer (100) or other processing means, a motion
controller
(102), the motion platform (I04), and an instrumented foot specimen (106). The
computer
-16
CA 02406622 2002-10-23
WO 01/85402 PCT/CA01/00678
(100) may be used both to execute the transformation and inverse kinematics
routines and
to collect and process the resulting force data.
Since the motion platform mimics the ground striking the foot, a computer must
execute software that manipulates and transforms the foot relative to ground
trajectory into
an appropriate ground relative to foot trajectory following the calculations
outlined above.
This software can be further used to shift and/or rotate this trajectory into
a position and/or
orientation that best exploits the motion platform's range of motion. These
translations or
rotations must, of course, be reflected in the positioning of the foot
specimen in the motion
platform's coordinate system. Inverse kinematics routines within the computer
software
can then be used to convert this desired ground trajectory into a set of six
leg trajectories
for the motion platform. These leg trajectories can then be formatted into
appropriate
motion commands and uploaded to the platform's motion controller.
Any industrial motion controller capable of regulating and coordinating the
position
of six axes can execute the motion program using either standard PID
(Proportional-
Integral-Derivative) or custom (e.g. adaptive) control algoritluns. The motion
controller
issues instruction to the motors' drives in turn, producing motion in each
motor axis.
If the motion platform trajectory has been appropriately planned, the strike
plate of
the platform will contact the shoe specimen. If the shoe has been insti-
umented with a six
degree of freedom force sensor, the forces of contact at the shoe can be
transformed into
forces in the ankle and returned for further processing as required. A screen
shot of
temporally distributed forces measured using an apparatus and method of the
present
invention is shown in Figure 10.
As will be apparent to those skilled in the art, various modifications,
adaptations and
variations of the foregoing specific disclosure can be made without departing
from the scope
of the invention claimed herein.
-17
