comparison of human postural response to ship motion encountered at sea and simulated motion in a...

40
` Comparison of Human Postural Response to Ship Motion Encountered at Sea and Simulated Motion in a Lab by Mohammed Thouseeq M. Eng. Project submitted to Faculty of Graduate Studies and Postdoctoral Affairs in partial fulfillment of the requirements for the degree of Master of Engineering in Mechanical and Aerospace Engineering Carleton University Ottawa, Ontario, Canada March 2014

Upload: mohammed-thouseeq

Post on 15-Aug-2015

30 views

Category:

Engineering


2 download

TRANSCRIPT

Page 1: Comparison of Human Postural Response to Ship Motion Encountered at Sea and Simulated Motion in a Lab

`

Comparison of Human Postural Response to Ship Motion

Encountered at Sea and Simulated Motion in a Lab

by

Mohammed Thouseeq

M. Eng. Project submitted to

Faculty of Graduate Studies and Postdoctoral Affairs in partial

fulfillment of the requirements for the degree of

Master of Engineering

in

Mechanical and Aerospace Engineering

Carleton University

Ottawa, Ontario, Canada

March 2014

Page 2: Comparison of Human Postural Response to Ship Motion Encountered at Sea and Simulated Motion in a Lab

`

i

Abstract

This report presents a comparison of the data collected from human subjects trying to

maintain their postural stability as they perform a secondary task when they are subjected to real

and simulated ship motion at sea and in the laboratory, respectively. The secondary task

consisted of identifying and transcribing odd numbers on a touch pad as well as on a

conventional writing pad. A number of factors contribute to the challenge of maintaining

postural stability, which may result in impairment of the certain functionalities of the human

body due to continuous motion. Motion sickness (MS) and motion induced fatigue are examples

of such impairment. MS can occur when a person is exposed to continuous motion; it can in turn

affect their performance and cause fatigue, loss of balance, and motivation. These problems

inhibit the subject from working effectively on a ship. The project included using a mathematical

model of the dynamics of the human body that calculates the angular motion that various joints

experience and the associated mechanical work performed by the subject while trying to

maintain their postural stability on a moving platform. The mathematical model consists of 15

body segments and 14 body joints resulting in ninety six degrees of freedom. The required data

were obtained using a 6 DOF motion platform system, a motion capture system, a load cell, Tek-

scan insole sensors, a Crossbow inertial measurement sensor, Xsens MTi sensor, and a GoPro

camera. Four subjects participated. A secondary task was performed by each subject, during

which they were asked to write on a touch pad as well as over a paper on a writing pad with the

platform in motion. The data were collected, processed, and compared with the corresponding

data for the same task on a ship during an experimental trial. The subjects were oriented at three

Page 3: Comparison of Human Postural Response to Ship Motion Encountered at Sea and Simulated Motion in a Lab

`

ii

different angles with each of the subjects facing six different conditions 0, 45, and 90 degrees

and using both tablet and paper recording media.

A comparison of the angular displacement of various joints at sea compared to those in

the laboratory is provided and analyzed.

Acknowledgements

I am very thankful to Dr. Fred Afagh and Dr. Robert Langlois for having me to do my

M. Eng. Project with them and for supporting my work throughout this project. My thanks also

go to everyone in the Applied Dynamics Laboratory and to the subjects for volunteering their

time and help throughout the project. I would like to thank especially Aren, Nick, and Heather. I

am particularly grateful to Gurwinder Kaur for providing me with guidance from the very

beginning of this project. I also thank Praveen Pullattu Jose and Fahd Basheer for helping me

with the experiment and data collection.

Page 4: Comparison of Human Postural Response to Ship Motion Encountered at Sea and Simulated Motion in a Lab

`

iii

TABLE OF CONTENTS

Abstract ............................................................................................................................................ i

Acknowledgements ......................................................................................................................... ii

TABLE OF CONTENTS ............................................................................................................... iii

1. Introduction ................................................................................................................................. 1

2. Literature Review: ...................................................................................................................... 3

2.1 Introduction of human postural stability and control ............................................................ 3

2.2 Postural stability .................................................................................................................... 4

2.3 Impact of motion on postural stability and task performance ............................................... 5

2.4 Impact of motion on performance of cognitive tasks ............................................................ 7

3. Experimental setup...................................................................................................................... 8

3.1 Modules of experimental setup ............................................................................................. 8

3.1.1 MOOG-6DOF Stewart Platform Motion Simulator ....................................................... 9

3.1.2 Opti-track Motion Capture System ............................................................................... 10

3.1.3 Tek-scan insole sensor system for transient foot pressure data .................................... 12

3.1.4 Crossbow sensor ........................................................................................................... 13

3.1.5 Xsens-MTi sensor ......................................................................................................... 14

3.1.6 Load cell ....................................................................................................................... 15

3.1.7 GoPro camera ............................................................................................................... 15

3.2 Data processing ................................................................................................................... 16

4. Joint angle calculation and full body matrix approach ............................................................. 17

4.1 Joint angle ........................................................................................................................... 17

Page 5: Comparison of Human Postural Response to Ship Motion Encountered at Sea and Simulated Motion in a Lab

`

iv

4.2 Full body matrix model ....................................................................................................... 19

5. Results ....................................................................................................................................... 21

5.1 The dominant angle at a joint .............................................................................................. 21

5.2 Comparison of the data from the sea trial with the laboratory results ................................ 23

5.3 Mechanical work performed by individual joints in combined ship motion ...................... 26

6. Conclusion ................................................................................................................................ 28

Appendix A- Comparison of ship and laboratory data of the subjects 2 and 3. ........................... 31

Appendix B - RMS matlab code ................................................................................................... 34

TABLE OF FIGURES

Figure 1 : An inverted model with the rotation along the ankle joint [8] ....................................... 6

Figure 2 : Experimental set-up with MOOG Stewart platform and railings [9]. ............................ 9

Figure 3 : Markers placed at 15 different body segments ............................................................. 10

Figure 4 : Opti-track picture of the skeleton captured during the T-pose. .................................... 11

Figure 5 : A foot pressure sensing Tek scan system with insoles 960 sensels. ............................ 12

Figure 6 : Crossbow AHRS 400 Sensor ....................................................................................... 13

Figure 7 : Xsens MTi sensor fixed over the helmet ...................................................................... 14

Figure 8 : Load cell ....................................................................................................................... 15

Figure 9 : GoPro camera placed over the helmet .......................................................................... 15

Figure 10 : Hip joint angle calculation between pelvis and right thigh [16] ................................ 18

Figure 11 : Axis of rotation at left knee [16] ................................................................................ 18

Figure 12 : Comparison of laboratory data with ship data – subject 1 ......................................... 24

Figure 13 : Mechanical work done at 0 , 45 , and 90 headings .................................................. 27

Figure 14 : Comparison of laboratory data with ship data – subject 2 ......................................... 31

Figure 15 : Comparison of laboratory data with ship data – subject 3 ......................................... 32

Figure 16 : Comparison of extension, abduction, and axial rotation angles ................................. 33

Page 6: Comparison of Human Postural Response to Ship Motion Encountered at Sea and Simulated Motion in a Lab

`

1

1. Introduction

Postural control has been defined as the control of the body’s position in space for the

purposes of balance and orientation [1]. Posture of the human body can be either in a stable or

unstable position. When on a moving platform such as a ship, there are a wide variety of reasons

one should be considering in order to keep the body stable such as environmental factors, motion

sickness, severity of sea state, visual effects, stance width, etc. When a human body is in postural

instability it often is preceded with motion sickness which is a common by-product of exposure

to optical depictions of inertial motion [2].

Human posture is mainly controlled by the central nervous system. During posture

maintenance it is the vision that gives a signal to the brain which in turn provides sufficient

movement to the body part to remain stable. Movements are mostly seen in the body joints at the

lower extremity, mainly at the ankles, knees, and hips. Sometimes even hands could be used to

hold the body or to lean towards an external support. Crew members normally need to perform

tasks at sea states in a maritime environment which can be of long duration. When the crew

members work on a moving platform the location of the centre of mass (CoM) of their body and

the centre of pressure (CoP) of their stance will change. It has been observed that the magnitude

of forces used in postural control were greater at sea than on land [3]. One of the major factors

that contribute highly to postural stability is the stance width. It has been found that as stance

width increases, postural stability increases thus decreasing the MS [4].

In this project the motion platform in the laboratory was subjected to angular motion profiles

replicating those recorded on the Canadian Forces Auxiliary Vessel Quest(CFAV) [5] during the

Page 7: Comparison of Human Postural Response to Ship Motion Encountered at Sea and Simulated Motion in a Lab

`

2

Q-348 Sea Trial. The data collected included position data of various body segments, foot

pressure data, metabolic energy data, and secondary task data. The angular motion between the

two adjacent segments at various joints was calculated and the results from the ship and platform

motion were compared. The secondary task data were also recorded and was kept for further

processing to obtain the effects of motion induced interruption on the performance of a particular

task.

Page 8: Comparison of Human Postural Response to Ship Motion Encountered at Sea and Simulated Motion in a Lab

`

3

2. Literature Review:

2.1 Introduction of human postural stability and control

Postural stability is one of the major factors that one should consider when dealing with

the efficiency of work performance, when the work is to be carried out in a moving environment.

To have better performance efficiency one has to attain postural stability. Postural stability can

be defined as the balanced body posture when the body segments are oriented relative to the

gravitational vector [6]. One attains postural stability by having proper control over the muscles

and by adjusting the angle of different joints after perceiving the environmental condition like

the type of environmental motion or inclination of the foot with the supporting base. The existing

literature indicates a significant volume of research regarding the general effects of

environmental motion on human performance like motion sickness, simulator sickness, balance

problems, physical fatigue, etc. [7]. These effects could sometimes affect the task performance.

The central nervous system, stance width, vision and somatosensory senses, vestibular

senses, etc. control the postural stability of the human system. Modern approaches to understand

postural control assume some sort of central processing of sensory information to produce body

reactions to external and internal disturbances and thus they resemble sensorimotor feedback

schemes [8].

Page 9: Comparison of Human Postural Response to Ship Motion Encountered at Sea and Simulated Motion in a Lab

`

4

2.2 Postural stability

The orientation of the human body in space is generally identified with respect to three

planes, i.e., the transverse plane, the coronal plane, and the sagittal plane [9]. As noted earlier,

different body segments and joints need to experience displacement in order to maintain stability.

This is achieved by the central nervous system with the coordination of the following three

subsystems - sensory, processing and motor subsystems.

The sensory system comprises of three components - visual, vestibular, and

somatosensory. These components in turn control the centre of mass (CoM) in coordination with

the central nervous system. CoM can be defined as a point equivalent of the total body mass in

the global reference system where the weighted average CoM of each body segment in 3D space

acts.[6]. As we know, the perturbation is felt at each of the body segments during a motion.

These perturbations are sensed by the somatosensory subsystem while the relative changes in the

external environment to the body position are sensed by the visual system. With the help of the

eyes, the vision system is able to detect the head position and orientation with respect to the

surroundings. The vestibular receptors sense the head angular velocity and the resultant of the

head translational and gravity accelerations which would be further processed by the processing

system [8].

After processing the different signals by the processing system the motor system plays a

further role. The motor system consists of muscles that actuate different joints and these muscles

are controlled by the central nervous system with the help of the motor neurons that are present

within each muscle. In general, the main body postural action takes place at the ankles in the so-

called ankle strategy, thus leading to an inverted-pendulum setup. Sometimes the body gets

stabilized by the hip strategy depending upon the motion the lower extremities (feet, legs, etc.)

Page 10: Comparison of Human Postural Response to Ship Motion Encountered at Sea and Simulated Motion in a Lab

`

5

experience [8]. Often depending upon the severity of the motion, the ankle strategy alone will

not be able to provide stability and a combined hip and ankle strategy will be necessary.

2.3 Impact of motion on postural stability and task performance

There exists significant literature on the study of the effects of the motion induced

perturbation of the position of various joints and segments of the body. Some results indicated

that motion primarily reduces the motivation to perform a specific task due to motion sickness,

and it increases fatigue due to increased energy requirements which in turn creates balance

problems [7].

Sometimes the perturbations are so large that the CoM will not be able to remain within the

base of the support area and one has to change the support area in order to maintain balance of

the body [10]. During such a process the stance width will not remain constant, and it would

need to be adjusted with respect to the motion in order to attain stability. With changes in stance

width, the effects of motion induced interruption and motion induced sickness vary. Riccio and

Stoffregen concluded in some of their research that the environmental motion could lead to

temporary instabilities in control of the movement, in general, and of bodily posture and

orientation, in particular [4]. They proposed that motion sickness would follow the development

of such instabilities in postural control and that motion sickness would occur only among persons

who exhibited postural instability.

Figure 1 depicts a human model representing a single inverted-pendulum of weight W which

is exposed to a motion in the sagittal plane by providing rotation around the ankle joint [8].

Page 11: Comparison of Human Postural Response to Ship Motion Encountered at Sea and Simulated Motion in a Lab

`

6

A B

In the figure, the position of the platform at A is horizontal and stationary, where the body

centre of mass is inclined at an angle α from the vertical axis. While in B the platform is tilting

with an angle ө about a vertical axis passing through the ankle joint making the inclination angle

of the line passing through the body center of mass to be ѱ = α+ө. Eventually when the motion

of the platform becomes severe the body will lose balance and be displaced, leading to motion

induced interruption. This would lead to a temporary delay of any task performed [11].

Figure 1 : An inverted model with the rotation along the ankle joint [8]

Page 12: Comparison of Human Postural Response to Ship Motion Encountered at Sea and Simulated Motion in a Lab

`

7

2.4 Impact of motion on performance of cognitive tasks

A cognitive task can be defined as a task that uses some memory or small mental work

for a short period of time. There exists considerable literature also about the impact of the

motion on cognitive tasks where various task have been considered. Some of the tasks included

memory comparison, or the use of pencil and paper [7]. In some other studies task duration was

extended to find if this affected the results, but most of the time the outcome was found to be

uneffected [12].

But if the feet alone is not able to maintain the body in a balance state, there are chances

of severe disturbances during the task performance. During a severe motion when the movement

of the lower extremeties alone is not able to maintain the body in a stable posture it will try to

move the upper extremeties. Under these conditions, if the task involves use of the hands, there

is a possibility of the task being interrupted for a short duration of time.

Page 13: Comparison of Human Postural Response to Ship Motion Encountered at Sea and Simulated Motion in a Lab

`

8

3. Experimental Setup

Introduction:

The objective of this experiment was to collect data from human subjects performing a task

while they try to maintain their postural stability on a moving platform. The subjects were placed

in three different orientations with respect to the simulated ship centreline, i.e. at 0 , 45 , and

90 . For each heading they were subjected to six different motions. A subject is comprised of 15

segment. The motion of each segment was captured using the markers placed over each

segments. The collected data were used to run an inverse dynamic Matlab code to calculate the

angular motion at each joint and the associated forces and moments.

3.1 Modules of experimental setup

The experimental setup consisted of the following modules:

1. MOOG-6 DOF Stewart platform motion simulator;

2. Opti-track motion capture system;

3. Tek-scan insole sensor system for transient foot pressure data;

4. Crossbow sensor;

5. Xsens MTi sensor;

6. Load cell with force plate;

7. Gopro camera.

Page 14: Comparison of Human Postural Response to Ship Motion Encountered at Sea and Simulated Motion in a Lab

`

9

Figure 2 : Experimental set-up with MOOG Stewart platform and railings [9].

3.1.1 MOOG-6DOF Stewart Platform Motion Simulator

A typical ship motion representing different sea states was simulated with the help of a Stewart

motion platform. The Stewart platform consists of 6 electro-mechanical actuators which provide

the platform with a 6 degree of freedom (DOF) synergistic mechanism. The controlled linear

motion of the actuators results in pitch, roll, yaw, heave, surge, and lateral motion of the

platform. Actuators are controlled by a motion base computer (MBC) which runs the control

software. Safe control of the motion base is maintained by the MBC which monitors the motor

position data, thermal switches, and amplifier faults. The command to the motion platform is

provided with the help of a host communication through the Applied Dynamics computer. With

the help of commands to the MBC, the motion platform could be at parked, engaged, or return to

its home position.

The input data for the Applied Dynamics computer was in .csv file format which was

provided within the 60 Hz limit. In this project each, subject was subjected to the motion with

same angular accelerations which they experienced during their Q-348 Sea Trial. The

Page 15: Comparison of Human Postural Response to Ship Motion Encountered at Sea and Simulated Motion in a Lab

`

10

translational acceleration was not been able to be reproduced in the lab as in the sea trial due to

the limited stroke of the motion base actuators. In table 1, the orientation of each subject at

different points in the data collection is shown.

Table 1: List of sea state profile

State 1 2 3 4 5 6

Orientation 0 45 90 0 45 90

3.1.2 Opti-track Motion Capture System

An Opti-track system consists of 8 cameras that are used to detect the motion of the human

body with the help of markers placed over the velcro suit worn by the subject. The Opti-track

system that was used in this experiment had 34 retro-reflective markers placed at different points

in the body parts. The placing of each marker was done in a manner to have reduced skin sliding

effects. The markers were placed at specifically defined positions over the 15 different body

segments [9].

Figure 3 : Markers placed at 15 different body segments

Page 16: Comparison of Human Postural Response to Ship Motion Encountered at Sea and Simulated Motion in a Lab

`

11

Calibration of the Opti-track system was done before collecting the data. There are two types of

calibrations, calibration of the cameras and calibration of the T-pose representing the subject’s

skeleton. For the camera calibration, the wanding procedure was used in which a single marker

wand handle is moved within the space inside the railing where the human subject would be

moving during the platform motion [13]. This procedure allowed calibrating the capturing space

and determining quality of the camera configuration and calibration. The highest quality is

achieved with 8 excellent qualities as output from all 8 cameras. Only one or two cameras with

lesser quality were acceptable. The next procedure was to set a ground plane and to define the

origin of the global frame of reference with the respective x, y, and z directions. In the skeleton

calibration procedure a virtual skeleton of the human subject was created. Here the human

subject wearing a Velcro suit with 34 markers over it was asked to stand in T pose, which was

done for 15 seconds and saved.

Figure 4 : Opti-track picture of the skeleton captured during the T-pose.

Page 17: Comparison of Human Postural Response to Ship Motion Encountered at Sea and Simulated Motion in a Lab

`

12

After the calibration procedures, the next step was to collect the data for a given motion to the

platform for 5 minutes. These data were then trajectorised and converted to .pt3 format which

were then smoothened after filling the gaps. For further analysis these data were saved as .c3d

files.

3.1.3 Tek-scan insole sensor system for transient foot pressure data

A Tek-Scan foot pressure system is a sensor used to collect the pressure produced by the

feet in the form of images with different colors. A Tek-Scan system consists of a pair of insoles

that is placed inside the shoes of the subject. An insole is made up of silver based links that are

arranged in 60 columns and 21 rows and embedded in a Mylar coating [14]. These columns and

rows intersect and produce 960 cells called sensels [14].

The calibration of the sensor was done by having the subject stand over one foot at a time

and interchanging the foot after 5 to 10 second. It was done twice by starting with the left foot

first and then repeating the same procedure with the right foot first for the second time. Each

calibration was done and saved under left and right foot files.

Figure 5 : A foot pressure sensing Tek scan system with insoles 960 sensels.

Page 18: Comparison of Human Postural Response to Ship Motion Encountered at Sea and Simulated Motion in a Lab

`

13

The pressure applied by the feet produce resistance in each of the cells. Thus with the

movement of the feet to maintain postural stability, the pressure exerted by the feet is changed

and this affects the resistance in the cells. The data from the cells were collected by the wire

which was pre-amplified and sent to the F-scan software in the computer that produced pressure

contour diagrams with different colours denoting the respective pressure levels. The color ranges

through red, blue, orange, yellow, and green. In the process of maintaining stability, depending

upon the motion severity, the maximum pressure applied by the feet can thus be detected.

3.1.4 Crossbow sensor

A Crossbow AHRS400 sensor was used in the experiment. It was kept over the platform

where the human subject was asked to stand. A sophisticated Kalman filter algorithm is used to

allow the unit to track orientation accurately through dynamic manoeuvres. It consists of

different sensors, which make use of the Coriolis force to detect angular rates independently of

accelerations.

Figure 6 : Crossbow AHRS 400 Sensor

Page 19: Comparison of Human Postural Response to Ship Motion Encountered at Sea and Simulated Motion in a Lab

`

14

The Crossbow sensor was placed at a corner on the motion platform surface to provide the pitch,

roll, and yaw data during the motion. It also provides combined linear accelerometers and

rotational rate measurements along the x, y, and z global co-ordinate axes during the motion in

the form of .csv files.

3.1.5 Xsens-MTi sensor

The Xsens-MTi is an inertial sensor which can measure angles, angular velocities, and

linear accelerations similar to the human vestibular system that provides information regarding

the movement and balancing done by the body parts. The MTi is a miniature, gyro‐enhanced

Attitude and Heading Reference System (AHRS). The orientation of the MTi is computed by the

Xsens Kalman (XKF-3) Filter for 3 degrees‐of‐freedom orientation. The XKF-3 algorithm works

as a sensor fusion algorithm where the measurement of gravity (by the 3D accelerometers) and

Earth magnetic north (by the 3D magnetometers) compensate for otherwise slowly, but

unlimited, increasing (drift) errors from the integration of rate of turn data (angular velocity from

the rate gyros) [15].

Figure 7 : Xsens MTi sensor fixed over the helmet

Page 20: Comparison of Human Postural Response to Ship Motion Encountered at Sea and Simulated Motion in a Lab

`

15

3.1.6 Load cell

The load cell used in this project consisted of a force plate to measure the reaction

forces and moment. It has 6 degrees of freedom which collects the three forces: Fgrfx; Fgrfy; Fgrfz

and three moments: Mgrfx; Mgrfy; Mgrfz developed under the left foot [9]. These forces and

moments were used in the calculation of the angular displacement and the work done at each

joint using the mathematical model that is discussed in a later section.

3.1.7 GoPro camera

Figure 8 : Load cell

Figure 9 : GoPro camera placed over the helmet

Page 21: Comparison of Human Postural Response to Ship Motion Encountered at Sea and Simulated Motion in a Lab

`

16

The GoPro is a durable camera that was installed on the subject’s helmet in order to

record high definition video of the view seen by each subject during the cognitive task

performance. Figure 9 shows the GoPro camera. Even though the data were not processed

specifically for this project, it could be used to determine the time taken to resume normal

posture after undergoing motion induced interruption and also the significant effect of vision on

human postural stability in the future. It could also be used to determine the delay in resuming

the secondary task after moiton-induced interruption.

3.2 Data processing

The laboratory data were collected for four individuals who had participated in the Q-348

Quest sea trial. At first, the data from the Opti-track system stored as .pt2 was trajectorized to

.pt3 which is then smoothened by filling the gaps present. For gaps less than 5 measurement

points in duration cubic spline interpolation is applicable, for gaps with 5 frames or greater the

rigid body angle interpolation is recommended [9]. These gaps were due to the Opti-track

cameras inability to capture the movement of markers that were covered by the presence of

railings on the platform.

Finally, the data received from the load cell and .c3d file was used as input to the

inverse dynamic Matlab code that uses the full matrix approach to find the angular displacements

and the work done at different body joints. The load cell data was saved in .csv format which

was trimmed from the point of a maximum value which was due to the stamping of the left foot

at the start of motion. The Rms value of the angular displacement at each of the body joints were

then calculated by using the Rms Matlab code.

Page 22: Comparison of Human Postural Response to Ship Motion Encountered at Sea and Simulated Motion in a Lab

`

17

4. Joint angle calculation and full body matrix approach

The full body matrix approach was used to find the angle subtended and the work done

by each of the body joints and segments [9]. For this we had to transform the data from the

global frame of reference to the local reference frame at each joint by rotational transformation.

The rotational transformation of the data was done as the captured data were in the global

reference frame. Three markers that were used on each of the body segments defined a local

coordinate system in three dimensional space. The resulting translation matrix that was used for

coordinate transformation is fully described in [9].

4.1 Joint angle

The human body is divided into 14 different joints and 15 body segments. The angle

produced by each of the joints is calculated for a particular instant of time by using the data

collected from the Opti-track system. For this the local coordinates of each of the body segments

were considerd. The local joint coordinate system is used to describe the angular displacement at

different body joints. This joint coordinate system was proposed by Chao (1980) and Grood and

Suntay (1983) for different body joints [9]. Three rotations are considered to take place between

two body segments. Figure 10 depicts these angles as an example in the case of the joint at the

right hip. Here we are considering the coordinate systems of the pelvis segment and the right

thigh segment adjacent to the hip joint [9, 16].

Page 23: Comparison of Human Postural Response to Ship Motion Encountered at Sea and Simulated Motion in a Lab

`

18

Here distal refers to the longitudinal axis and proximal the medio-lateral axis and the floating axis is

defined as:-

Floating Axis =

| |

There are three angles at the joint that are expressed as:

α = flexion/extension angle;

β = abduction/adduction angle; and

γ = axial rotation angle

Here α is the rotation of the proximal segment about the medio-lateral axis (x-axis), β is the

rotation about the floating axis (z-axis), and γ (y-axis) is the rotation about the longitudinal axis.

For the case of the right hip joint, these angles are calculated as follows;

Figure 10 : Hip joint angle calculation between pelvis and right thigh [16]

Figure 11 : Axis of rotation at left knee [16]

Page 24: Comparison of Human Postural Response to Ship Motion Encountered at Sea and Simulated Motion in a Lab

`

19

α = 90 – arccos ( proximal joint)

β = –90 – arccos ( proximal distsl)

γ = 90 – arccos ( distal joint)

To find the corresponding angles at the remaining body joints the above procedure was used

by considering the different body segments adjacent to the body joints. A prewritten inverse

dynamic Matlab code by Kaur [9] was used in order to measure these angles whose Rms values

were determined later.

4.2 Full body matrix model

In the full body matrix model, the dynamics of the human body motion is cast as a ninety

degrees of freedom system representing 96 unknowns, i.e., force components and moment

components associated with each of the 16 joints between the segments [9]. This requires the

development of a 90×90 coefficient matrix, [A]. The matrix entries are calculated using the

Newton Euler equations of motion for each segment. The ground force reaction and the moment

component of the left foot are used as the known values of the system. The remaining 90

unknown force and moment components [X] of the 15 body segment joints are then determined

using the following matrix equation:

[X] 90×1 = [A]-1

90×90 [B] 90×1 (1)

Here [B] is a column vector representing the known inertia terms associated with each segment.

Using the Newton-Euler equation of motion, for each segment, the forces and moments in the

local coordinate system of each segment become:

Page 25: Comparison of Human Postural Response to Ship Motion Encountered at Sea and Simulated Motion in a Lab

`

20

Fx = ˗ ˗

(2)

Fy = ˗ ˗

(3)

Fz = ˗ ˗

(4)

where, Fx, Fy, and Fz are reaction force components acting in x, y, and z directions representing

the medio-lateral, longitudinal, and anterior-posterior directions respectively; ,

,

are translational acceleration components in the three respective directions; and ,

,

are the three ground reaction force components; and mgx, mgy, and mgz define the weight vector

in three-dimensional space [9].

The rotational dynamics equation are given as;

Mx = Ixxαx + (Izz ˗ Iyy)ωzωy ˗ ˗ ˗ (5)

My= Iyyαy + (Ixx ˗ Izz)ωxωz ˗ ˗ ˗ (6)

Mz = Izzαz + (Iyy ˗ Ixx)ωyωx ˗ ˗ ˗ (7)

In the above equations Ixx, Iyy, and Izz are principal components of the inertia matrix while αx, αy,

and αz represent components of segment angular acceleration along the x, y, and z axes. For a

single body segment, equations (2) through (7) are cast as a 6×1 matrix with the first three rows

representing the forces and the bottom three representing the moments, respectively,

Page 26: Comparison of Human Postural Response to Ship Motion Encountered at Sea and Simulated Motion in a Lab

`

21

[

]

[

]

=

[

-

-

- ]

On applying the inverse matrix solution the above matrices are rearranged as follows:

[

]

=

[

]

[

-

-

- ]

Thus following the above procedure for all 15 segments a 90×90 coefficient matrix is formed

which is used to obtain the unknown forces and moments. After finding the unknown forces and

moments, the work done by each joint is calculated. The following equation is used in order to

find the work done by one of the body joints by moving from an angle ;

= ∫

The total work done by the body joint is calculated by finding the summation of the work done at

each of the 16 body joints.

5. Results

5.1 The dominant angle at a joint

Depending on the particular joint, one of the three angular motions, i.e. flexion angle,

abduction angle , or axial rotation angle, reflected the dominant motion at the respective joint.

Page 27: Comparison of Human Postural Response to Ship Motion Encountered at Sea and Simulated Motion in a Lab

`

22

Table 2 : The dominant angle used by 14 joints of the subject

For example at some joints the flexion angle dominates over the abduction angle and the axial

rotation angle. Therefore, the dominant angle at each joint was identified and its motion under

various conditions was studied and compared. Figure 16 in Appendix A shows the column chart

of the three angles moved by the different body joints of a typical subject. The results indicate

that the angular motions of each of the 14 joints of the human subject were dominated by one of

the angles.

Table 1 shows the dominant angle at each of the 14 joints of the body during the motion in

the laboratory. Here α (flexion/extension angle), β (abduction/adduction angle), and γ (axial

rotation angle) are the rotation of the proximal segment about the medio-lateral axis (x-axis), the

floating axis (z-axis), and the longitudinal axis respectively. It can be observed in Figure 16 in

Appendix A that the axis of rotation by most of the joints when the subjects were oriented

differently remained to be consistent at varying severities of motion.

JOINT DOMINANT ANGLE

head β

left ankle α

right ankle α

left elbow α

right elbow α

left hip γ

right hip γ

left knee γ

right knee γ

left shoulder γ

right shoulder γ

left wrist β

right wrist β

pelvis β

Page 28: Comparison of Human Postural Response to Ship Motion Encountered at Sea and Simulated Motion in a Lab

`

23

Only in case of the right elbow was there a slight change in the dominant rotation axis as the

severity of the motion increased. The axial rotation angle had the dominant motion in slower

platform motions while the flexion angle became the dominant angle in the case of severe

motions. This might have been due to the missing data from the Opti-track files. For most of the

upper extremity joints such as the head, and the right and the left wrists, the dominant motion

was about the floating axis. In case of the right and left ankles the dominant movement was

about the medio-lateral axis. For majority of the joints the dominant rotation was around their

longitudinal axis. But for more severe platform motions the motions in the ankles were decreased

while the motions in the hip joints were increased.

5.2 Comparison of the data from the sea trial with the laboratory results

Figure 11 shows the comparison of the root mean square value of the dominant angle

experienced at each joint for subject 1 during the ship trial vs. the corresponding values for the

platform motion in the laboratory. In most cases the angular motions experienced at each joint

are higher during the laboratory experiments than the corresponding values during the sea trial.

At sea states 1 and 4 when the subject was facing at 0o with respect to the ship heading, only 5 of

the 14 joints experienced higher angles in the sea trial than at the laboratory. For sea states 2 and

6 at 45o orientation

only 4 of the joints experienced larger motion on the ship than in the lab.

While at sea state 6 with subject facing 90o

the number of such joint was 10.Due to the fact that

translational acceleration and large displacement amplitudes were present while at sea and not in

the lab, it was hypothesized that larger amplitude joint motion would have been measured at sea.

But from the results it doesn’t seems to support the fact as we should have observed lower

motion amplitudes in the lab due to the absence of translated accelerations and other factors from

the sea.

Page 29: Comparison of Human Postural Response to Ship Motion Encountered at Sea and Simulated Motion in a Lab

`

24

State 1-0 degree State 2-45 degree

State 3-90 degree State 4-0 degree

State 5-45 degree State 6-90 degree

Figure 12 : Comparison of laboratory data with ship data - subject 1

Page 30: Comparison of Human Postural Response to Ship Motion Encountered at Sea and Simulated Motion in a Lab

`

25

There are different possible reasons that might have led to the above results. On looking

into the output data from the inverse dynamic Matlab code, it could be found that there were a

large number of missing data. The reason behind such missing data might have been due to three

reasons. At some point during the motion the cameras on the Opti-track system could not have

been able to detect the position of the markers placed on the subject. This was due to the fact that

obstruction created by the railings, required for safety, caused positioning of the body such that

one segment superpose the markers of another segment. It is possible that these missing data

might have the larger values which could have resulted in the higher angle on the ship. But in

some of the cases, like the result in state 5, even the joint angles that had larger values (sea trial

in this case) still had missing data. It could be due to the time interval at which these readings

were missing had only fewer movement in the joint while the remaining readings were quite high

to provide the larger readings.

Finally the last conceivable reason could be the influence of the eye sight on the sea trial

and laboratory. A study, by A. J Weins, explained the effect of motion on the subject with eyes

open and closed [17]. It was seen that the head angle had a larger level of variability during eyes

open when compared with the eyes closed while motion induced interference were found with

eyes closed. The large variability had been due to the change in centre of mass over which the

head was aligning during stable posture. But when the eyes were closed, the visual system

couldn’t send signals to the central nervous system to realign the head along the axis of the

centre of mass that led to lesser motion by the head. Thus we could conclude that the vision has

an effect over the movement. During the motion of the ship during the sea trial the wall faced by

the subject was moving in the same direction along with them. But this was not the same in the

Page 31: Comparison of Human Postural Response to Ship Motion Encountered at Sea and Simulated Motion in a Lab

`

26

laboratory, as the platform was moving, the wall faced by subjects was stationary creating a sort

of opposite movement to the eyes. So it is possible to have higher perturbations perceived by the

mind as well as the motion induced interruption would have affected the subject during the

simulated motions.

5.3 Mechanical work performed by individual joints in combined ship motion

The work done by 14 different joints are plotted in 6 different states with varying motion

severity. Figure 13 shows the pie chart of the mechanical work done by the 4 different subjects

after finding their average work done. Looking into the total mechanical work done by the ankle

joint, as the severity in motion increases, there is a decrease in percentage of work done. At state

1 work done by the ankle joint is 39% which gradually reduces to 29% in state 2, while there was

an increase in total mechanical work done by the knee from 19% to 42% in the respective states.

The least percent of work done by the ankle was found in the state 6 which is about 17% where

the motion was found the most severe out of all the motions.

It could be concluded that most of the work was done by the ankle when the motion was

not severe but gradually decreased and got divided by the other joints including both the upper

extremities and the lower extremities. The hip angle work done was found to be quite varying

from 4% to 21% which was decreasing and increasing irregularly at different states.

Page 32: Comparison of Human Postural Response to Ship Motion Encountered at Sea and Simulated Motion in a Lab

`

27

State 1-0 degree

State 6-90 degree

State 3-90 degree State 4-0 degree

State 2-45 degree

`State 5-45

degree

Figure 13 : Mechanical work done at 0 , 45 , and 90 headings

.

Page 33: Comparison of Human Postural Response to Ship Motion Encountered at Sea and Simulated Motion in a Lab

`

28

So it was able to find both the ankle and hip strategy in most of the motions. Here L5-S1

is the work done by joints in lumbar-sacral region, which had a range of 1% to 9% contribution

to the work. Considering the upper extremities mechanical work done by most joints where

pretty much lower like the joint at the shoulders, wrists and elbows just accounted for 1% to 9%.

It might be due to their position of holding the touch pad or writing pad for the secondary task at

the same place from the beginning to the end of the motion. There was a large movement in the

head joint, which varied from 6% to 56%. Thus the contribution of different joints could be seen

to vary with motion with maximum work done by the head that might have resulted from the

visual interference with the body being in motion while the wall was stationary as discussed

earlier.

6. Conclusion

The comparison of the data from the ship and lab using the joint angles was done in this

project. Throughout the experiment the ankle and hip strategy had been seen during the work

done by the joints to maintain stable posture. Results have provided information regarding the

extent of body motions experienced at sea and in the lab as well as the participation of the

different joints of the body. It shows the importance to provide similar moving environment in

the surroundings to the subject in the lab by using some sort of screen that can avoid visual

differences. The secondary task data was not analyzed as it was not available. As a future work,

the above result can be made more meaningful by removing the particular time period from the

lab during which the data was found missing in the ship. Finding out the Rms after refining the

data could lead us closer to the reason behind such deviation in the present results or even getting

a more meaningful comparison.

Page 34: Comparison of Human Postural Response to Ship Motion Encountered at Sea and Simulated Motion in a Lab

`

29

References

[1] M. Woollacott and A. Shumway-Cook, "Attention and the control of posture and gait:

a review of an emerging area of research," Gait and Posture, vol. 16, pp. 1-14, 2002.

[2] L. Smart, T. Stoffregen and B. Bardy, "Visually Induced Motion Sickness Predicted

by Postural Instability," Human Factors: The Journal of Human Factors and Ergonomics

Society, vol. 44, pp. 451-465, 2002.

[3] T. A. Stoffregen, S. Villard, F. Chen and Y. Yu, "Standing Posture on Land and at

Sea," Ecological Psychology, vol. 23, pp. 19-36, January-March, 2011.

[4] T. A. Stoffregen, K. Yoshida, S. Villard, L. Scibora and B. G. Bardy, "Stance Width

Influences Postural Stability and Motion Sickness," Ecological Psychology, vol. 22, pp.

169-191, July-September, 2010.

[5] N. Bourgeois, R. Langlois, and A. Hunter, "Quest Q-348 Sea Trial: Human Postural

Stability Studies," Unpublished journal, Carleton University.

[6] P. D A Winter PhD, "Human balance and posture control during standing and

walking," vol. 3, pp. 193-214, 1995.

[7] A. H. WERTHEIM, "Working in a moving environment," Ergonomics, vol. 41, pp.

1845-1858, December 1998, 1998.

[8] K. A. Tahboub, "Biologically-inspired humanoid postural control," Journal of

Physiology - Paris, vol. 103, pp. 195-210, 200909, 2009.

[9] G. Kaur, "Mechanical Energy Expenditure while Maintaining Postural Stability in

Shipboard Motion Environments," Carleton University, 2013.

[10] C. A. Duncan, "Biomechanical Adaptations Required to Maintain Postural Stability

in Moving Environments when Performing Manual Materials Handling Activities,"

University of New Brunswick, Canada, 2007.

[11] P. Crosslandl and K. Rich, "VALIDATING A MODEL OF THE EFFECTS OF

SIDP MOTION ON POSTURAL STABILITY," 1998.

[12] W. Bles, "Experiments on Motion Sickness Aboard the MV Zeefakkel," Technische

Universiteit Delft, Faculteit der Werktuigbouwkunde en Maritieme Techniek, Vakgroep

Scheepshydromechanica, 1991.

[13]Calibration, "http://www.naturalpoint.com/optitrack/products/arena/tutorials.html".

Page 35: Comparison of Human Postural Response to Ship Motion Encountered at Sea and Simulated Motion in a Lab

`

30

[14] A. L. Randolph, M. Nelson, S. Akkapeddi, A. Levin and R. Alexandrescu,

"Reliability of measurements of pressures applied on the foot during walking by a

computerized insole sensor system," Arch. Phys. Med. Rehabil., vol. 81, pp. 573-578,

2000.

[15] X. MTi and M. U. Manual, "Technical Documentation," Product Manual.Xsens Co,

pp. 2-30, 2006.

[16] C. L. Vaughan, B. L. Davis and J. C. O'connor, "Dynamics of Human Gait," Human

Kinetics Publishers Champaign, Illinois, 1992.

[17] A. J. Weins, "Ship deck postural stabilty and joint angles," 2010.

Page 36: Comparison of Human Postural Response to Ship Motion Encountered at Sea and Simulated Motion in a Lab

`

31

State 1-0 degree State 2-45 degree

State 3-90 degree State 4-0 degree

State 5-45 degree State 6-90 degree

Appendix A- Comparison of ship and laboratory data of the

subjects 2 and 3.

`

Figure 14 : Comparison of laboratory data with ship data – subject 2

Page 37: Comparison of Human Postural Response to Ship Motion Encountered at Sea and Simulated Motion in a Lab

`

32

State 1-0 degree State 2-45 degree

State 3-90 degree State 4-0 degree

State 5-45 degree State 6-90 degree

Figure 15 : Comparison of laboratory data with ship data – subject 3

Page 38: Comparison of Human Postural Response to Ship Motion Encountered at Sea and Simulated Motion in a Lab

`

33

State 2-45 degree

Figure 16 : Comparison of extension, abduction, and axial rotation angles

State 1-0 degree

State 5-45 degree State 6-90 degree

State 4-0 degree State 3-90 degree

Page 39: Comparison of Human Postural Response to Ship Motion Encountered at Sea and Simulated Motion in a Lab

`

34

Appendix B - RMS Matlab code

Code to find the root mean square value of the joint angles

clc;

clear all;

%each of the 14 joint alpha angle imported%

alpha _angle = importdata('C:\Users\mohammedthouseeq\Desktop\ ship output\ ');

% removing the NaN values from each of the rows

alpha_ angle(isnan(alpha_ angle(1,:))) = [];

%rms value of each joint

rms_alpha_ angle=rms(alpha_ angle);

%each of the 14 joint beta angle imported%

beta_ angle = importdata('C:\Users\mohammedthouseeq\Desktop\ ship output\');

% removing the NaN values from each of the rows

beta_ angle(isnan(beta_ angle(1,:))) = [];

%rms value of each joint

rms_beta _angle=rms(beta_ angle);

%each of the 14 joint gamma angle imported%

gamma_ angle = importdata('C:\Users\mohammedthouseeq\Desktop\ ship output\’);

% removing the NaN values from each of the rows

gamma_ angle(isnan(gamma_ angle(1,:))) = [];

%rms value of each joint

rms_gammaangle=rms(gamma_ angle);

Page 40: Comparison of Human Postural Response to Ship Motion Encountered at Sea and Simulated Motion in a Lab

`

35

%final matrix containing three angles

rms=[rms_alpha_head_angle,rms_beta_head_angle,rms_gamma_head_angle;

rms_alpha_leftankle_angle,rms_beta_leftankle_angle,rms_gamma_leftankle_angle;

rms_alpha_rightankle_angle,rms_beta_rightankle_angle,rms_gamma_rightankle_angle;

rms_alpha_leftelbow_angle,rms_beta_leftelbow_angle,rms_gamma_leftelbow_angle;

rms_alpha_rightelbow_angle,rms_beta_rightelbow_angle,rms_gamma_rightelbow_angle;

rms_alpha_lefthip_angle,rms_beta_lefthip_angle,rms_gamma_lefthip_angle;

rms_alpha_righthip_angle,rms_beta_righthip_angle,rms_gamma_righthip_angle;

rms_alpha_leftknee_angle,rms_beta_leftknee_angle,rms_gamma_leftknee_angle;

rms_alpha_rightknee_angle,rms_beta_rightknee_angle,rms_gamma_rightknee_angle;

rms_alpha_leftshoulder_angle,rms_beta_leftshoulder_angle,rms_gamma_leftshoulder_angle;

rms_alpha_rightshoulder_angle,ms_beta_rightshoulder_angle,rms_gamma_rightshoulder_angle;

rms_alpha_leftwrist_angle,rms_beta_leftwrist_angle,rms_gamma_leftwrist_angle;

rms_alpha_rightwrist_angle,rms_beta_rightwrist_angle,rms_gamma_rightwrist_angle;

rms_alpha_pelvis_angle,rms_beta_pelvis_angle,rms_gamma_pelvis_angle];

%matrix containig maximum value of the row

rms_max=max(rms,[],2);