Work and fatigue characteristics of unsuited and suited humans during isolated isokinetic joint motions

L. JAVIER GONZALEZ†, J. C. MAIDA§∗, E.H. MILES‡

S.L. RAJULU†† and A.K. PANDYA‡

†SPACEHAB, Houston, Texas

§ (P.I.) NASA Johnson Space Center, Houston, Texas

‡Lockheed Martin Space Operations, Houston, Texas

††NSBRI, Houston, Texas

Keywords: Joint fatigue; Joint strength; Task planning; Extra-Vehicular Activity (EVA);

Extra-vehicular Mobility Unit (EMU)

The effects of a pressurized suit on human performance were investigated. The suit is

known as an Extra-vehicular Mobility Unit (EMU) and is worn by astronauts while

working outside of their space craft in low earth orbit. Isolated isokinetic joint torques of

three female and three male subjects (all experienced users of the suit in 1G) were

measured while working at 100% and 80% of their maximum voluntary torque (MVT),

(here MVT is synonymous with maximum voluntary contraction (MVC)). It was found

that the average decrease in the total amount of work done when the subjects were

wearing the EMU was 48% and 41% while working at 100% and 80% MVT,

respectively. There is a clear relationship between the MVT and the time and amount of

∗ Author for correspondence. E-mail: jim.maida@jsc.nasa.gov

1

work done until fatigue. In general the stronger joints took longer to fatigue and did

more work than the weaker joints. It was found that the EMU decreases the work output

at the wrist and shoulder joints the most. This is due to the EMU joint geometry. The

average total amount of work done by the test subjects increased by 5.2% (20.4%) for the

unsuited (suited) case, when the test subjects decreased the level of effort from 100% to

80% MVT. Also, the average time to fatigue increased by 9.2% (25.6%) for the unsuited

(suited) case, when the test subjects decreased the level of effort from 100% to 80%

MVT. The EMU also decreased the joint range of motion. It was also found that the

experimentally measured torque decay could be predicted by a logarithmic equation. The

absolute average error in the predictions was found to be 18.3% and 18.9% for the

unsuited and suited subject, respectively, working at 100% MVT, and 22.5% and 18.8%

for the unsuited and suited subject, respectively, working at 80% MVT. These results

could be very useful in the design of future EMU suits, and planning of Extra-Vehicular

Activity (EVA) for the upcoming International Space Station assembly operations.

1. Introduction

With the upcoming International Space Station assembly missions, EVA will increase

dramatically. Approximately 900 EVA hours will be required to assemble the Space

Station with an additional 200 hours per year for maintenance requirements. Task

planning could increase the efficiency of the suited astronaut while performing Extra-

Vehicular Activities (EVAs). Efficient task planning could be facilitated by a simulation

using a three-dimensional human model with proper range of motion (ROM), strength,

and fatigue characteristics for all the major joints of the body.

2

While no comprehensive database exists which contains strength and fatigue

measurements for all the isolated joints of an unsuited and suited astronaut, much

research has been done to investigate the ergonomic characteristics of unsuited and suited

astronauts (O’Hara 1989; Shafer et al. 1992; Bishu and Klute 1993; Poliner et al. 1993,

1994; Rajulu and Klute 1993; Rajulu et al. 1993; Bishu et al. 1994; Wilmington et al.

1994; Morgan et al. 1996, Rajulu et al. 1998). One objective of this study was to acquire

and analyze experimental data for unsuited and suited humans while performing isolated

joint motions for all the major joints of the upper body (the upper body joints are the ones

most commonly used to perform EVAs). It is desirable, in particular, to study the

differences in the time to fatigue and total work done on a joint-by-joint basis, for suited

and unsuited humans, in order to compare and evaluate suit designs. An additional long-

term objective is to develop a predictive model for work and fatigue (Pandya et al.

1992a). This requires identifying and quantifying a trend in the torque decay over time

for each joint axis direction. Quantifying this torque decay is an objective of the present

research. The torque decay trend in conjunction with a maximum available torque model

(Pandya et al. 1992b) can be used to predict work and fatigue. This model could then be

integrated into a model for simulation and optimization of tasks in suited and unsuited

conditions.

2. Experiments

2.1 Subjects

Three female and three male subjects were used in the experiments. The low subject

count was due to time constraints. First, the availability of a suit and a qualified subject

3

at the same point in time was limited. Second, time spent wearing the suit was a

constraint, being generally two to three hours for one session, thus creating the need for

many sessions per subject with the additional session set up time. Third, because we

were measuring fatigue, additional time per subject was required in order to reach the

50% MVT threshold and to allow for subject recovery. Finally, the subjects participating

in the experiments were required to have a current Air Force Class 3 physical. In order to

minimize learning effects, only subjects with considerable experience in the use of the

EMU were allowed to participate in the experiments. The experience in the EMU for all

subjects is 41 hours. Table 1 gives a description of the test subjects.

[Insert table 1 about here]

2.2 Test facility and equipment

The experiments were conducted in the Precision Air Bearing Facility (PABF) at Johnson

Space Center. This facility allowed for multiple test configurations through the use of its

floor sled, smooth floor area, and air bearings. Extremely heavy objects mounted on top

of air bearings can easily be moved around the PABF floor once the air supply to the

bearings has been activated. This aided in positioning the test hardware and subjects in

proper alignment for conducting the strength and fatigue tests.

Joint strength and fatigue measurements were made using a LIDO Multi-Joint II

testing unit (Loredan Biomedical, Inc., West Sacramento, CA). The LIDO Multi-Joint II

is an integrated system consisting of a dynamometer connected to a personal computer.

The system has a series of attachments that enable strength measurements of the various

joints of the human body. The system also allows setting of experimentally defined

4

exercise parameters such as: velocity, ROM, and torque limits. A unique feature of the

LIDO is that it incorporates a gravity compensation feature that takes into account the

one-gravity artifact throughout the ROM. This capability is used to remove the effect of

the weight of the LIDO attachment, EMU suit segment, and the subject’s limbs from the

measured torques and forces, providing the subject’s actual effort. The gravity

compensation routine works as follows: the subject passively grasps the attachment as it

travels through the ROM, during which the LIDO computer calculates the torque

generated by the weight of the attachment, body segment, plus suit segment (when

applicable) in order to subtract these artifacts from subsequent measurements.

Standard Shuttle EMU suits with external air supplies were used for these

experiments. This suit is known to restrict the motion at the shoulder and wrist joints.

The scye opening, which is the opening through which the arm exits, is much larger than

the shoulder circumference, while the suit wrist joint is located two to three inches above

the actual wrist joint. An environmental control system was also used to maintain the suit

pressurization at 29.6 kPa. During testing, a floor sled was used to relieve the test subject

of the mass of the suit (approximately 118 to 123 kg). Figure 1 shows a suited subject

during one of the tests. The unsuited tests were also done using the floor sled in the

PABF.

[insert figure 1 about here]

2.3 Procedure

Since upper-body movements (e.g., locomotion using hand rails, using tools to repair

objects, and moving large masses) are integral to most EVA operations, emphasis was

5

placed on testing strength and fatigue for the major joints of the upper body. The five

isolated isokinetic joint motions that were included were: (1) shoulder flexion/extension,

(2) shoulder abduction/adduction, (3) shoulder internal/external rotation, (4) elbow

flexion/extension, and (5) wrist flexion/extension. All of the testing was performed on

the subject’s right side with the subject secured in an upright (standing) posture.

The procedure to find the MVT is now described. The subject drove the LIDO

attachment back and forth using maximum exertion at a prescribed angular velocity of

1.05 rad/s throughout the ROM for three non-stop repetitions. Because of time

constraints only one angular velocity was used. The angular velocity of 1.05 rad/s was

selected as the benchmark based on previous work done by Pandya et al. 1992b. The

variation between the three torque curves was then computed using the LIDO computer.

A variation of more than ten percent was used as a measure to indicate that the subject

was not using maximal exertion during the strength test. If the variation was less than ten

percent, the subject’s MVT for both directions of motion was computed by averaging the

three peak torque values for each direction.

For the fatigue tests the subject drove the LIDO attachment repeatedly at the

prescribed angular velocity of 1.05 rad/s throughout the entire ROM using the target level

of exertion (100% or 80% of the MVT) without stopping. During the submaximal

fatigue tests (80% of the MVT), the subject received real-time visual feedback of her/his

performance plotted against the target submaximal force level on the data acquisition

computer display. For the first repetition, the subject was given a window of ± 3% of the

target level. The subject was allowed to continue the test until the torque output in both

directions dropped below the defined fatigue index value (50% of the MVT) for three

6

consecutive repetitions. If the subject did not reach fatigue after five minutes, the test

was stopped. A five-minute rest period was allowed between joint measurements. This

procedure was used for both the unsuited and suited tests. The 80% level of exertion was

determined from pilot testing, since it provided the best range of data before reaching the

50% of the MVT mark.

The use of 50% of the MVT as the fatigue index requires some explanation. This

number came from the work of Patton et al. 1978, which found that the use of 50% MVT

as a fatigue index ensures a significant decline in function. Others have also used this

protocol to denote fatigue, such as Schwendner et al. 1995 who concluded that an

isokinetic fatigue test to 50% MVT, repeated three times, was an appropriate fatigue

generating protocol for most active males.

3. Results

3.1 Unsuited versus suited performance

The time to fatigue presented here was computed in a manner different than the protocol

used in the experiments. Here, the time to fatigue is defined as the ending time of the

repetition for which the computed work done during that repetition dropped below 50%

of the work done during the first repetition. This method was used because it was desired

to quantify fatigue as a function of the decline in energy output, and we believe that this

is a more overall measure of fatigue. The time to fatigue was then used as the ending

time for computing the number of repetitions, and work to fatigue. Figure 2 illustrates

how the time to fatigue was computed as just described for subject 5 during elbow

7

flexion while working at 100% of the MVT. For the case shown in figure 2, subject 5

fatigued at 73 seconds for the unsuited test, and at 53 seconds for the suited test. Figure 3

shows the mean torque per repetition, for subject 5 during elbow flexion while working at

100% of the MVT, while figure 4 shows polar plots of the torque measurements, also for

subject 5 during elbow flexion while working at 100% of the MVT.

[insert figures 2, 3 and 4 about here]

Tables 2 and 3 show the MVT, number of repetitions, time, and work to fatigue

(for each joint motion averaged across all subjects) for the unsuited and suited subject

working at 100% and 80% of their MVT, respectively. Table 4 shows the percent

decrease in work done by the test subjects from unsuited to suited for each joint motion.

Figures 5 and 6 show bar plots of the work done to fatigue for each joint motion averaged

across all subjects working at 100% and 80% MVT, respectively.

[insert tables 2, 3, and 4 about here]

[insert figures 5 and 6 around here]

3.2 Prediction of torque decay as a function of time

The procedure to obtain equations to predict the torque decay as a function of time for a

specific joint motion is now described. First, the mean torque for each repetition and

corresponding mean time was computed for the fatigue task. The data for all the subjects

was then grouped and smoothed before it was normalized. A moving average of ten

consecutive data points was visually found to provide sufficient smoothing of the data

while keeping the characteristics of the torque decay. The data was normalized using the

8

maximum torque value of the grouped smoothed data. It was then found, by trial and

error (with the use of the solver function in Microsoft Excel), that an equation of the

form:

| , (1) ct|ln ba norm ++=τ

where τnorm is the normalized torque, t is time, and a, b and c are coefficients, represents

the torque decay quite well. Figure 7 shows the actual normalized data and curve fits for

shoulder flexion for the unsuited and suited subjects working at 100% of their MVT.

Table 5 gives the coefficients for equation (1) for unsuited and suited subjects working at

100% of their MVT, while table 6 gives the coefficients for equation (1) for unsuited and

suited subjects working at 80% of their MVT. Given equation (1) and the maximum

torque (τmax) for the first repetition of a repetitive task, the torque decay (τ) as a function

of time can then be predicted by:

|]ct|ln b[a max ++=ττ . (2)

Using equation (2) the predictions were computed for subjects 1-6. The average absolute

error in the predictions for all subjects was 18.3% for the unsuited case and 18.9% for the

suited case at 100% MVT, and 22.5% for the unsuited case and 18.8% for the suited case

at 80% MVT. Table 7 gives the absolute average error in the predictions for each joint

averaged across the six subjects. Figure 8 shows the actual and predicted torque for

subject 2 during wrist extension at 100% MVT, while figure 9 shows the actual and

predicted torque for subject 3 during shoulder internal rotation at 80% MVT.

[insert figure 7 about here]

[insert tables 5 and 6 about here]

[insert figures 8 and 9 about here]

9

4. Discussion

One objective of this study was to quantify the effects of an EMU suit on the human

performance. Another objective was to develop equations that can be used to predict the

torque decay for each joint motion tested. In this section, a brief discussion of the

justification of the experimental and modeling procedures is provided. This is followed

by a discussion of the results pertaining to the two objectives mentioned above.

4.1 Justification of assumptions

For the data analysis, the female and male data were grouped. In light of the fact that

individuals use the same muscle groups, and that these muscle groups basically have the

same shape, orientation, and points of attachment, but differ in magnitude of force

exerted (Pandya et al. 1992b), it is felt that grouping of the data is justified since it was

normalized. For this reason, the low number of test subjects used is also thought to be

sufficient to give the trends of the work and fatigue characteristics of unsuited and suited

humans during isolated isokinetic joint motions.

4.2 Unsuited versus suited performance

The effect of the suit on performance is clearly evident. In general, the work and mean

torque per repetition decreased when the subject was suited (figures 2 and 3,

respectively). The joint range of motion also decreased in general for the suited subject

(figure 4). The average decrease in the total amount of work done when the subjects

were wearing a suit was 48% and 41% while working at 100% and 80% MVT,

10

respectively. There is a clear relationship between the MVT and the amount of work

done until fatigue. In general, the larger the MVT the more work that was done (tables 2

and 3). Further, for the case with the test subjects working at 100% MVT, wrist

extension and shoulder flexion are affected the most, while at 80% MVT, wrist flexion

and shoulder external rotation are affected the most (table 4). As mentioned previously,

the EMU is known to restrict the motion at the shoulder and wrist joints; therefore, these

findings are in check with the known restrictions of the EMU.

The average total amount of work done by the test subjects increased by 5.2%

(20.4%) for the unsuited (suited) case, when the test subjects decreased the level of effort

from 100% to 80% MVT, this corresponds to an increase of work done by the test

subjects of 62 Joules (128 Joules) for the unsuited (suited). Also, the average time to

fatigue increased by 9.2% (25.6%) for the unsuited (suited) case, when the test subjects

decreased the level of effort from 100% to 80%MVT, this corresponds to an increase of

time to fatigue for the test subjects of 10 seconds (20 seconds) for the unsuited (suited).

These results require some explanation. While working at 80% of the MVT the suit

offers less resistance, therefore, the effort of the subjects is directed towards moving the

Lido attachment rather than working against the suit. It makes sense then, that the time to

fatigue when working at 80% MVT increases, and consequently so does the amount of

work done. The significance of these results is obvious, i.e., astronauts should be

instructed to work at an MVT less then 100% MVT if the time to complete the task is not

an issue. These results could be very useful in the planning of EVAs.

4.3 Prediction of torque decay as a function of time

11

It was found that the torque decay could be predicted by a logarithmic equation (equation

(2), figures 7, 8 and 9). The average correlation coefficient, R2, for the curve fit for all

the joint motions is 0.89 for both the unsuited and suited cases at 100% MVT, and 0.91

and 0.89 for the unsuited and suited cases, respectively, at 80 % MVT, which indicates a

good fit. The absolute average error in the predictions was found to be 18.3% and 18.9%

for the unsuited and suited subject, respectively, working at 100% MVT, and 22.5% and

18.8% for the unsuited and suited subject, respectively, working at 80% MVT. As

mentioned in the introduction, a torque decay trend, used in conjunction with a maximum

available torque model (Pandya et al. 1992b), can be used to predict work and fatigue.

This predictor can also be integrated into a three-dimensional dynamic model for

analyses of tasks in both suited and unsuited conditions.

4.4 General discussion

The main intent of this research was to quantify the fatigue effects of wearing a

pressurized suit and to develop predictor equations for the torque decay as a function of

time. Pandya et al. 1992b, developed relations to predict dynamic isolated joint torques

from an individual’s lean body mass, however, in the work presented here, the lean body

mass of the subjects was not recorded since the intent of this study was not to relate

fatigue to lean body mass. With this in mind, it should be noted that the subjects in these

experiments belong to a specific population, i.e., they are all healthy subjects and

regularly participate in fitness training programs, and therefore, findings from this

research study may not be generalizable. But, these findings could be generalized to

specific skill tasks that require fit individuals.

12

5. Summary and future work

The effects of an EMU on isolated joint torque were measured. Using six subjects, both

suited and unsuited, isolated isokinetic joint torques were measured with the subjects

working at 100% and 80% of their MVT, using 50% of the MVT as the fatigue index. It

was found, on average, that the EMU reduced total work by about 50%, and the larger the

MVT, the longer the time to fatigue and the greater amount of work done. It was found

that the EMU decreases the work output at the wrist and shoulder joints the most. This is

due to the EMU joint geometry as mentioned earlier. The amount of work and time to

fatigue improved the most for the suited case, when the subject started at 80% MVT. The

EMU also decreased the joint range of motion. It was also found that the experimentally

measured torque decay trend could be predicted by a logarithmic equation with an

average error of ± 9.3% and 10.3% for the 100% and 80% MVT cases, respectively.

These results could be very useful in the design of future EMU suits, and planning of

Extra-Vehicular Activity (EVA\) for International Space Station assembly missions.

Future work includes the incorporation of these results into a human three-dimensional

dynamic model.

±

6. Acknowledgements

This project was funded by NASA grant number 96-HEDS-05.

13

References

BISHU, R.R. and KLUTE, G. 1993, Investigation of the effects of extravehicular activity

(EVA) gloves on performance, NASA Technical Paper 3401.

BISHU, R.R., BRONKEMA, L.A., GARCIA, D., KLUTE, G. and RAJULU, S. 1994,

Tactility as a function of grasp force: Effects of glove, orientation, pressure, load,

and handle, NASA Technical Paper 3474.

MORGAN, D.A., WILMINGTON, R.P., PANDYA, A.K., MAIDA, J.C. and DEMEL,

K.J. 1996, Comparison of extravehicular mobility unit (EMU) suited and unsuited

isolated joint strength measurements, NASA Technical Report TP-3613.

O’HARA, J.M. 1989, The effect of pressure suit gloves on hand performance,

Proceedings of the Human Factors Society 33rd Annual Meeting, 139-142.

PANDYA, A.K., MAIDA, J.C., ALDRIDGE, A.M., HASSON, S.M. and WOOLFORD,

B.J. 1992a, The validation of a human force model to predict dynamic forces

resulting from multi-joint motions, NASA Technical Paper 3206.

14

PANDYA, A.K., HASSON, S.M., ALDRIDGE, A.M. and MAIDA, J.C. 1992b,

Correlation and prediction of dynamic human isolated joint strength from lean

body mass, NASA Technical Paper 3207.

PATTON, R.W., HINSON, M.M., ARNOLD, B.R. and LESSARD, B. 1978, Fatigue

curves of isokinetic contractions, Arch. Phys. Med. Rehabili., 59, 507-509.

POLINER, J., WILMINGTON, R.P., KLUTE, G.K. 1993, Strength capabilities and load

requirements while performing torquing tasks in zero gravity, NASA Technical

Paper 3433.

POLINER, J., WILMINGTON, R.P. and KLUTE, G.K. 1994, Geometry and gravity

influences on strength capability, NASA Technical Paper 3511.

RAJULU, S.L., KAKAVAND, A., MULANI, K. and MAIDA, J.C. 1998, Hand strength

in a simulated microgravity environment, 28th International Conference on

Environmental Systems, (Danvers, Massachusetts).

RAJULU, S.L. and KLUTE, G.K. 1993, A comparison of hand grasp breakaway

strengths and bare-handed grip strengths of the astronauts, SML III test subjects,

and the subjects from the general population, NASA Technical Paper 3286.

15

RAJULU, S.L., POLINER, J., KLUTE, G.K. 1993, Loads produced by a suited subject

performing tool tasks without the use of foot restraints, NASA Technical Paper

3424.

SCHAFER, L.E., RAJULU, S.L. and KLUTE, G.K. 1992, A comparison of two shuttle

launch and entry suits: reach envelope, isokinetic strength, and treadmill tests,

22nd International Conference on Environmental Systems, (Seattle, Washington).

SCHWENDNER, K.I., MIKESKY, A.E., WIGGLESWORTH, J.K. and BURR, D.B.

1995, Recovery of dynamic muscle function following isokinetic fatigue testing,

International Journal of Sports Medicine, 16(3), 185-189.

WILMINGTON, R.P., POLINER, J. and KLUTE, G.K. 1994, Use of a pitch adjustable

foot restraint system: Operator strength capability and load requirements, NASA

Technical Paper 3477.

16

Tables

Table 1. Description of test subjects.

Subject 1 Subject 2 Subject 3 Subject 4 Subject 5 Subject 6 Gender female male male male female female Mass (kg) 59 77 86 66 68 54 Height (cm) 165 172 181 170 165 168 Hand dominance right right left right right right EMU experience (hrs) 20 60 70 25 40 35

17

Tabl

e 2.

Rep

etiti

ons,

time

and

wor

k to

fatig

ue fo

r uns

uite

d an

d su

ited

subj

ects

wor

king

at 1

00%

of t

heir

MV

T (s

tand

ard

devi

atio

ns in

pare

nthe

sis)

.

U

NSU

ITE

D

SU

ITE

DJo

int m

otio

n |M

VT

| (N

⋅m)

Rep

etiti

ons

Tim

e (s

) W

ork

(J)

|MV

T| (

N⋅m

) R

epet

ition

s T

ime

(s)

Wor

k (J

) W

rist e

xten

sion

11

(4)

40 (1

7)

107

(47)

24

3 (7

3)

7 (2

) 25

(5)

59 (8

) 96

(35)

W

rist f

lexi

on

18 (7

) 22

(13)

61

(41)

23

6 (1

18)

11 (5

) 19

(7)

47 (1

8)

115

(93)

El

bow

ext

ensi

on

43 (1

5)

29 (9

) 11

0 (2

7)

1199

(613

) 34

(15)

30

(10)

83

(26)

76

9 (3

83)

Elbo

w fl

exio

n 39

(17)

18

(5)

69 (1

8)

711

(427

) 33

(16)

16

(5)

43 (9

) 37

7 (2

09)

Shou

lder

ext

ensi

on

63 (2

7)

23 (4

) 10

8 (1

8)

2017

(120

7)

67 (3

1)

27 (1

5)

91 (5

9)

1134

(597

) Sh

ould

er fl

exio

n 61

(16)

24

(6)

108

(26)

19

23 (4

32)

42 (1

2)

26 (8

) 87

(43)

66

8 (2

24)

Shou

lder

abd

uctio

n 50

(13)

45

(15)

20

2 (7

3)

2520

(729

) 34

(31)

56

(23)

15

2 (6

5)

1207

(185

) Sh

ould

er a

dduc

tion

54 (2

3)

28 (2

2)

125

(99)

16

60 (9

07)

41 (3

3)

41 (1

3)

112

(55)

11

07 (3

57)

Shou

lder

ext

erna

l 21

(8)

25 (8

) 89

(29)

43

6 (2

41)

19 (8

) 15

(7)

48 (1

8)

234

(205

) Sh

ould

er in

tern

al

39 (1

4)

30 (1

0)

108

(35)

10

49 (4

07)

37 (9

) 18

(3)

60 (1

2)

562

(103

) A

vera

ge

4028

109

1199

3327

7862

7

18

Tabl

e 3.

Rep

etiti

ons,

time

and

wor

k to

fatig

ue fo

r uns

uite

d an

d su

ited

subj

ects

wor

king

at 8

0% o

f the

ir M

VT

(sta

ndar

d de

viat

ions

in

pare

nthe

sis)

.

U

NSU

ITE

D

SU

ITE

DJo

int m

otio

n |M

VT

| (N

⋅m)

Rep

etiti

ons

Tim

e (s

) W

ork

(J)

|MV

T| (

N⋅m

) R

epet

ition

s T

ime

(s)

Wor

k (J

) W

rist e

xten

sion

11

(4)

34 (1

6)

95 (4

7)

228

(61)

7

(2)

41 (2

2)

105

(69)

14

0 (5

2)

Wris

t fle

xion

18

(7)

30 (1

0)

79 (2

8)

324

(104

) 11

(5)

26 (1

7)

63 (4

3)

99 (7

8)

Elbo

w e

xten

sion

43

(15)

34

(15)

14

9 (7

0)

1193

(502

) 34

(15)

37

(15)

95

(31)

10

05 (6

04)

Elbo

w fl

exio

n 39

(17)

27

(15)

11

1 (6

7)

890

(568

) 33

(16)

22

(10)

58

(23)

51

4 (2

76)

Shou

lder

ext

ensi

on

63 (2

7)

32 (1

0)

148

(47)

24

26 (9

38)

67 (3

1)

31 (1

2)

97 (3

9)

1257

(105

4)

Shou

lder

flex

ion

61 (1

6)

19 (4

) 84

(20)

14

47 (3

97)

42 (1

2)

28 (7

) 83

(17)

77

5 (2

04)

Shou

lder

abd

uctio

n 50

(13)

35

(4)

144

(19)

23

27 (8

8)

34 (3

1)

69 (3

7)

152

(57)

13

10 (6

10)

Shou

lder

add

uctio

n 54

(23)

39

(20)

17

3 (7

9)

2095

(105

6)

41 (3

3)

62 (4

4)

148

(96)

13

28 (9

57)

Shou

lder

ext

erna

l 21

(8)

32 (1

2)

114

(49)

45

3 (1

21)

19 (8

) 22

(14)

71

(36)

27

2 (1

98)

Shou

lder

inte

rnal

39

(14)

28

(1)

98 (1

) 12

24 (1

84)

37 (9

) 31

(16)

10

8 (4

2)

851

(352

) A

vera

ge

4031

119

1261

3337

9875

5

19

Table 4. Percent decrease in work done to fatigue from unsuited to suited case.

Joint motion 100% MVT 80% MVT Wrist extension 60 39 Wrist flexion 51 69

Elbow extension 36 16 Elbow flexion 47 42

Shoulder extension 44 48 Shoulder flexion 65 46

Shoulder abduction 52 44 Shoulder adduction 33 37

Shoulder external rotation 46 40 Shoulder internal rotation 46 30

Average 48 41

20

Tabl

e 5.

Coe

ffic

ient

s for

equ

atio

n (1

) and

cor

rela

tion

coef

ficie

nt R

2 for 1

00%

MV

T.

U

nsui

ted

Su

ited

Join

t mot

ion

a b

c R

2

a

b c

R2

Wris

t ext

ensi

on

1.46

1-0

.192

11.0

610.

900.

917

-0.0

86-0

.385

0.77

Wris

t fle

xion

1.

268

-0.1

864.

211

0.96

1.34

5-0

.206

5.32

50.

95El

bow

ext

ensi

on

1.20

8-0

.122

5.50

20.

761.

568

-0.2

3311

.416

0.90

Elbo

w fl

exio

n

1.

350

-0.1

707.

815

0.81

1.13

2-0

.156

2.33

20.

81Sh

ould

er e

xten

sion

1.

486

-0.2

0011

.378

0.91

1.28

1-0

.154

6.22

20.

86Sh

ould

er fl

exio

n

1.

468

-0.1

8113

.367

0.93

1.34

5-0

.177

7.01

70.

94Sh

ould

er a

bduc

tion

1.85

8-0

.239

36.1

190.

891.

407

-0.1

7110

.742

0.94

Shou

lder

add

uctio

n

2.

257

-0.3

2249

.292

0.90

1.29

1-0

.164

5.92

40.

90Sh

ould

er e

xter

nal r

otat

ion

2.13

9 -0

.334

30

.202

0.

92

1.60

0 -0

.238

12

.412

0.

89

Shou

lder

inte

rnal

rota

tion

1.44

9 -0

.187

11

.006

0.

94

1.34

4 -0

.195

5.

813

0.94

21

Tabl

e 6.

Coe

ffic

ient

s for

equ

atio

n (1

) and

cor

rela

tion

coef

ficie

nt R

2 for 8

0% M

VT.

U

nsui

ted

Su

ited

Join

t mot

ion

a b

c R

2

a

b c

R2

Wris

t ext

ensi

on

1.37

4-0

.181

7.86

30.

870.

941

-0.0

67-0

.410

0.78

Wris

t fle

xion

2.

189

-0.3

0350

.797

0.80

1.33

2-0

.148

9.36

70.

80El

bow

ext

ensi

on

1.66

2-0

.223

19.5

340.

933.

061

-0.4

7278

.713

0.88

Elbo

w fl

exio

n

1.

663

-0.2

4914

.358

0.93

1.71

8-0

.281

12.9

290.

92Sh

ould

er e

xten

sion

1.

615

-0.2

1417

.757

0.93

2.34

9-0

.347

49.1

340.

94Sh

ould

er fl

exio

n

1.

427

-0.1

7511

.424

0.90

1.13

2-0

.115

3.14

30.

87Sh

ould

er a

bduc

tion

2.70

6-0

.372

98.3

010.

911.

231

-0.1

157.

454

0.90

Shou

lder

add

uctio

n

1.

534

-0.1

8318

.503

0.91

1.14

5-0

.116

3.50

00.

91Sh

ould

er e

xter

nal r

otat

ion

1.91

1 -0

.280

25

.835

0.

96

2.07

6 -0

.327

26

.807

0.

95

Shou

lder

inte

rnal

rota

tion

1.85

4 -0

.250

30

.285

0.

95

1.75

2 -0

.251

20

.007

0.

96

22

Tabl

e 7.

Abs

olut

e av

erag

e pe

rcen

t erro

r in

torq

ue p

redi

ctio

n fo

r eac

h jo

int.

Uns

uite

dSu

ited

Join

t mot

ion

100%

MV

T

80%

MV

T

100%

MV

T

80%

MV

T

Wris

t ext

ensi

on

18.2

23.0

15.1

25.5

Wris

t fle

xion

29

.730

.715

.419

.3El

bow

ext

ensi

on

12.6

22.3

17.2

19.0

Elbo

w fl

exio

n

14

.623

.924

.219

.4Sh

ould

er e

xten

sion

16

.416

.317

.316

.3Sh

ould

er fl

exio

n

15

.231

.123

.217

.6Sh

ould

er a

bduc

tion

11.6

18.9

16.4

17.2

Shou

lder

add

uctio

n

26

.324

.418

.218

.8Sh

ould

er e

xter

nal r

otat

ion

18.3

17

.8

25.7

21

.7

Shou

lder

inte

rnal

rota

tion

20.6

16

.2

16.4

13

.3

Ave

rage

18

.322

.518

.918

.8

23

FIGURE CAPTIONS

Figure 1. Suited subject with LIDO Multi-Joint II testing unit while being supported by air bearing floor sled.

Figure 2. Work per repetition for subject 5 during elbow flexion at 100% MVT. Figure 3. Mean torque per repetition for subject 5 during elbow flexion at 100% MVT. Figure 4. Polar plots of torque measurements for subject 5 during elbow flexion at 100%

MVT (angles in degrees). Figures 4c and 4d are close up views of 4a and 4b, respectively.

Figure 5. Work to fatigue for test subjects working at 100% MVT. Figure 6. Work to fatigue for test subjects working at 80% MVT. Figure 7. Normalized actual data and logarithmic curve fit for shoulder flexion at 100%

MVT. Figure 8. Actual and predicted torque for subject 2 during wrist extension at 100% MVT. Figure 9. Actual and predicted torque for subject 3 during shoulder internal rotation at

80% MVT.

24

25

0 50 100 150 200 250 3005

10

15

20

25

30

35

40

45

50 W

ork

Per R

ep (J

)

Time (s)

Unsuited Suited 50% of Work Done in 1st Rep (Unsuited) 50% of Work Done in 1st Rep (Suited)

26

0 50 100 150 200 250 3000

5

10

15

20

25

30

Mea

n To

rque

Per

Rep

(New

ton

met

ers)

Time (s)

Unsuited Suited

27

Unsuited Suited (a) (b)

(c) (d)

20

40

30

210

60

240

90

270

120

300

150

330

180 0

Joint Torque (Newton meters)

20

40

30

210

60

240

90

270

120

300

150

330

180 0

Joint Torque (Newton meters)

0

180

start time fatigue time start time fatigue time

28

0

500

1000

1500

2000

2500

3000

wrist e

xt

wrist fl

ex

elbow

ext

elbow

flex

shou

lder e

xt

shou

lder fl

ex

shou

lder a

b

shou

lder a

d

shou

lder e

xt

shou

lder in

t

Wor

k (J

)

UnsuitedSuited

29

0

500

1000

1500

2000

2500

3000

wrist e

xt

wrist fl

ex

elbow

ext

elbow

flex

shou

lder e

xt

shou

lder fl

ex

shou

lder a

b

shou

lder a

d

shou

lder e

xt

shou

lder in

t

Wor

k (J

)

UnsuitedSuited

30

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

0 50 100 150 200 250 300

Time (s)

Nor

mal

ized

Mea

n To

rque

Per

Rep

Unsuited (actual)

Suited (actual)

Unsuited (log fit)

Suited (log fit)

31

0

2

4

6

8

10

12

0 20 40 60 80 100 120

Time (s)

Mea

n To

rque

Per

Rep

(New

ton

met

ers)

Actual (unsuited)

Predicted (unsuited)

Actual (suited)

Predicted (suited)

32

-50

-45

-40

-35

-30

-25

-20

-15

-10

-5

0

0 20 40 60 80 100 120

Time (s)

Mea

n To

rque

Per

Rep

(New

ton

met

ers)

Actual (unsuited)Predicted (unsuited)Actual (suited)Predicted (suited)

.

33