Signature redacted-- Siqnature redacted

Evaluating Human-EVA Suit Injury Using

Wearable SensorsMASS

by

Ensign Sabrina Reyes, U.S. Navy

B.S., Aerospace EngineeringUnited States Naval Academy (2014)

Submitted to the Department of Aeronautics and Astronauticsin partial fulfillment of the requirements for the degree of

Master of Science in Aeronautics and Astronautics

at the

MASSACHUSETTS INSTITUTE OF TECHNOLOGY

ACUS0 ILNSTITUTE)F TECHNOLOGY

JUN 28 2016

IBRARIESARCHIVES

June 2016

@ Massachusetts Institute of Technology 2016.

A uthor ...............

Certified by..

All rights reserved.

Signature redacted;.......

Department of Aeronauticand Astronautics

V~ \ %, \ May 19, 2016

Signature redacted--. ...........

Jelfrey A. Hoffman, Ph.D.Professor of theractice, Aeronautics and Astronautics

Siqnature redactedA ccepted by .................. I........ ..............................

PauloI C Lozno7n PhDT

Associate Professor of Aeronautics and AstronauticsChair, Graduate Program Committee

am=

2

Evaluating Human-EVA Suit Injury Using Wearable Sensors

by

Ensign Sabrina Reyes, U.S. Navy

Submitted to the Department of Aeronautics and Astronauticson May 19, 2016, in partial fulfillment of the

requirements for the degree ofMaster of Science in Aeronautics and Astronautics

Abstract

All the current flown spacesuits are gas pressurized and require astronauts to exert asubstantial amount of energy in order to move the suit into a desired position. Thepressurization of the suit therefore limits human mobility, causes discomfort, andleads to a variety of contact and strain injuries. While suit-related injuries have beenobserved for many years and some basic countermeasures have been implemented,there is still a lack of understanding of how humans move within the spacesuit. Therise of wearable technologies is changing the paradigm of biomechanics and allowinga continuous monitoring of motion performance in fields like athletics or medical re-habilitation. Similarly, pressure sensors allow a sensing capability to better locatethe areas and magnitudes of contact between the human and their interface and re-duce the risk of injuries. Coupled together these sensors allow a better understandingof the complex interactions between the astronaut and his suit, enhance astronautsperformance through a real time monitoring and reducing the risk of injury. Thefirst set of objectives of this research are: to gain a greater understanding of thishuman-spacesuit interaction and potential for injury by analyzing the suit-inducedpressures against the body, to determine the validity of the particular sensors usedwith suggested alternatives, and to extend the wearable technology application toother relatable fields such as soldier armor and protective gear. An experiment wasconducted in conjunction with David Clark Incorporated Company on the LaunchEntry Development spacesuit analyzing the human-spacesuit system behavior for iso-lated and functional upper body movement tasks: elbow flexion/extension, shoulderflexion/extension, shoulder abduction/adduction and cross body reach, which is acomplex succession of critical motions for astronaut and pilot task. The contact pres-sure between the person and the spacesuit was measured by three low-pressure sen-sors (the Polipo) over the arm, and one high-pressure sensor located on the shoulder(Novel). The same sensors were used in a separate experiment conducted in con-junction with Protect the Force Company on several different United States MarineCorps (USMC) protective gear configurations, which analyzed the human-gear in-teractions for: shoulder flexion/extension, horizontal shoulder abduction/adduction,vertical shoulder abduction/adduction, and the cross body reach. Findings suggest

3

that as suit pressurization increases, contact pressure across the top of the shoulderincreases for all motion types. While it proved to be a perfectly acceptable methodfor gathering shoulder data, improvements can be made on the particular sensorsused and the type of data collected and analyzed. In the future, human-suit interfacedata can be utilized to influence future gas-pressurized spacesuit design. Addition-ally, this thesis briefly explores the incompatibilities between Russian and U.S. EVAcapabilities in order to make a case for equipment standardization.

Thesis Supervisor: Jeffrey A. Hoffman, Ph.D.Title: Professor of the Practice, Aeronautics and Astronautics

4

Acknowledgments

First and foremost, I would like to thank the wonderful advisors I had at MIT, without

whom this thesis could have never happened. To Dr. Jeff Hoffman, thank you for

all your honest guidance during the thesis process and regarding my aspirations to

become an astronaut. To Dr. Dava Newman, thank you for introducing me to

the world of human spaceflight and reigniting my passion for aerospace. My MIT

experience would not have been the same without such a wonderful person in my life

to give me incredible opportunities like meeting Buzz Aldrin, skiing in Montana for

a conference, or working with spacesuits and other fantastic people for my research.

Thank you both for all the opportunities and the unwavering support.

To the EVA team, Pierre, Alexandra, and Allie, I cannot thank you enough for all

your help on this thesis. It seriously would not have happened without you. Thanks

for all the fun meetings, for being some of my first friends at MIT, and for being

incredibly patient and helpful with all my questions even after you had moved on to

bigger and better things!

To Tony, John, and Grant, thank you guys for being my Navy partners in crime.

You guys understand and tolerate my awful mood swings, humor, and personality

probably more than anybody else, and for that I am so grateful. I am happy I had you

all to provide advice and/or sounding boards for weird Navy situations like P-codes,

disappearing without leave, etc. Tony and Grant, I guess I'm sort of happy we will

all be in the same pipeline so I can see your ugly faces even after we leave MIT, and

John, I am going to miss you so much but I know that you'll kick butt in flight school!

To Hannah, thanks for being the sweetest roommate, officemate, classmate, etc.

Our weird cookie binges, burger quests, lunch runs, and wonderful conversations kept

me from insanity (seriously). Thank you for being such a wonderful and patient

friend.

To Conor, Richard, Lynn, Forrest, Eddie, and the rest of the MVLers, you guys

are seriously the most amazing people ever! I am highly convinced that the Man

Vehicle Lab is the coolest, funnest lab at MIT, plus we produce some darn good

5

research. Thanks for letting me waste all your time because I don't feel like doing

any of my own work. Thanks for lab lunches, lab dog-sitting adventures, IEEE skiing,

HST formal shenanigans, Captain America movie nights, and all the other incredible

memories that I will cherish forever. I will miss you all so much, please come visit

me wherever I am in the Navy!

To the close friendships: Macauley, Anne, Parker, Patricia, Mark, and Emily,

thank you guys for random dinners, drink nights, and for distracting each other from

research and life. I love you guys to the moon and back.

A special thanks to Liz Zotos, Barb DeLaBarre, Ed Ballo, and Beth Marois for

providing advice, help, and friendship throughout my stay at MIT.

Finally, I would like to thank my family for their incredible prayers, love, support,

and encouragement. To Elizabeth, thank you for adopting my family into your own,

because you have become such an important part of our family. Thank you for all

your spot-on advice in all areas of life, because no one else seems to understand my

way of thinking quite like you do. I love all of you very much and would not have

gotten to where I am without you.

6

Contents

1 Introduction

2 Literature Review

2.1 Extravehicular Activity . . . . . . . . . . . . . . . . . . . . . . . . . .

2.2 EVA Training and Injury . . . . . . . . . . . . . . . . . . . . . . . . .

2.3 Previous Work on Development of a Quantitative Understanding of

Human-Spacesuit Interaction . . . . . . . . . . . . . . . . . . . . . . .

3 Sensor Systems and Experimental Design

3.1 Sensor System s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

3.1.1 Low Pressure Sensing System, the "Polipo" . . . . . . . . . ..

3.1.2 Novel High-Pressure Shoulder Sensor . . . . . . . . . . . . . .

3.1.3 APDM Inertial Measurment Units . . . . . . . . . . . . . . . .

3.2 Spacesuit Testing Experimental Design . . . . . . . . . . . . . . . . .

3.3 Marine Protective Gear Experimental Design . . . . . . . . . . . . . .

4 Novel System Results and Discussion

4.1 David Clark Experiment . . . . . . . . . . . .. . . . . .

4.1.1 Pressure Distributions . . . . . . . . . . . . . .

4.1.2 Pressure Profiles . . . . . . . . . . . . . . . . .

4.1.3 Statistical Analysis . . . . . . . . . . . . . . . .

4.2 Protect the Force Armor Gear Prototype Experiments

4.3 Conclusions and Future Work . . . . . . . . . . . . . .

7

13

16

16

19

23

25

25

25

28

29

30

32

36

. . . . . 36

. . . . . 37

. . . . . 41

. . . . . 47

. . . . . 53

. . . . . 55

5 International EVA Capabilities 58

5.1 A Case for EVA Standardization . . . . . . . . . . . . . . . . . . . . 58

6 Conclusions 66

A Human-Suit Interface Pressure Evaluation 68

8

List of Figures

2-1 Extravehicular Mobility Unit and Exploded View Diagram. (Image

Sources: NASA, Hamilton Sustrand) . . . . . . . . . . . . . . . . . . 17

2-2 Pivoted HUT on left and Planar HUT on right. Note the different

angles of the scye bearings in the two HUTs. (Image Source: NASA) 18

2-3 David Clark Launch and Entry Development Suit . . . . . . . . . . . 18

2-4 Astronaut training in the NBL in an inverted position (Image Source:

N A SA ) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20

3-1 Printed carbon-grease sensor with electrode extensions. (Image Source:

W yss at Harvard, 2014) . . . . . . . . . . . . . . . . . . . . . . . . . 27

3-2 AMOHR two-stranded conductive tape used for second Polipo iteration. 28

3-3 Experimental Sensor Systems: A) Low-pressure Polipo sensors, B)

High-pressure Novel shoulder sensor, C) APDM Opal inertial mea-

surement unit. (Image Source: Anderson, 2014) . . . . . . . . . . . . 29

3-4 Placement of the in-suit sensor systems. (Image Source: Anderson, 2014) 30

3-5 Descriptions of the four upper body motions performed during the

spacesuit experiment: three isolated joint motions (elbow flexion/extension,

shoulder flexion/extension, shoulder abduction/adduction), and one

functional task (cross body reach). (Image Source: Anderson, 2014,

Hilbert et al. 2014) . . . . . . . . . . . . . . . . . . . . . . . . . . ..31

9

3-6 Experimental Design Test Protocol: Each movement group consists

of a counterbalanced ordering of the motions. The motions studied

were: three isolated joint motions (elbow flexion/extension, shoulder

flexion/extension, and shoulder abduction/adduction) and a functional

task motion (cross body reach). In each movement group, the specific

motion was repeated 5 times for a total of 15 repetitions per motion. 32

3-7 Different USMC protection gear configurations used during testing . 33

3-8 Horizontal shoulder abduction/adduction. Beginning with their arms

extended at shoulder height, shoulder width apart, the subject bends

their arms at the shoulder in the transverse plane. The subject moves

through his or her maximum range of motion. The subject returns to

the initial start position then releases to the relaxed position. . . . . . 34

3-9 Protective Gear Experimental Design Test Protocol. Each movement

group consists of a counterbalanced ordering of the motions. The mo-

tions were: three isolated joint motions (shoulder flexion/extension,

shoulder abduction/adduction vertical, shoulder abduction/adduction

horizontal) and a functional task motion (cross body reach). In each

movement group, the specific motion was repeated 5 times for a total

of 15 repetitions per motion. . . . . . . . . . . . . . . . . . . . . . . . 35

4-1 Orientation of the Novel Sensor. The orientation corresponds to the

information in each pressure distribution figure. Coloring is for orien-

tation and is not related to pressure scales. (Image Source: Hilbert

20 15) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 7

4-2 Contact pressure distributions for each motion at each pressurized con-

dition for Subject 1. . . . . . . . . . . . . . . . . . . . . . . . . . . . 38

4-3 Contact pressure distributions for each motion at each pressurized con-


4-4 Contact pressure profiles for all motions in the 2.5-psi pressurized con-


10







4-8 Effects of pressurization on mean contact pressure for Subject 1. . . . 49

4-9 Effects of pressurization on mean contact pressure for Subject 2. . . . 49

4-10 Effects of motion type on mean contact pressure for Subject 1. . . . . 50

4-11 Effects of motion type on mean contact pressure for Subject 2. . . . . 50

4-12 Effect of subject on mean contact pressure for intermediate pressuriza-

tio n . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 1

4-13 Effect of subject on mean contact pressure for full pressurization. . . 51

4-14 Pressure distributions of all motions for different USMC armor config-

urations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54

4-15 Pressure distributions for rifle carry motion for Configuration 1 (on

left) and Configuration 4 (on right). . . . . . . . . . . . . . . . . . . . 55

4-16 Novel Single S2012 Sensor with 2 cm diameter (Image Source: novel.de) 57

5-1 Rear entry opening for Russian Orlan-M spacsuit. (Image Source:

NASA)........ ................................... 59

5-2 The U.S. EMU and Russian Orlan-M spacesuit shown side by side.

(Image Source: NASA) . . . . . . . . . . . . . . . . . . . . . . . . . . 63

11

List of Tables

3.1 Anthropometrics from typical Marine compared to anthropometrics of

subject used . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34

12

Chapter 1

Introduction

Human spaceflight programs are facing new challenges rising from the evolution of the

exploration agenda, the need for Commercial Crew and the new entry on the market

of space tourism. These different activities bring new challenges: planetary explo-

ration missions will require intensive extravehicular activities (EVA), space tourism

will require new, cheap and user friendly space systems, specifically, pressure suits.

Spacesuits need to adapt to this new era of space exploration and democratization

of space. Spacesuits are technical marvels: their main functions are providing oxy-

gen, pressure, food, water, waste removal, communication, thermal control, mobility,

radiation protection, direct sunlight protection, and micrometeorite protection. The

human body cannot survive in the vacuum of space because all air the air would

rush out of the lungs, blood vessels would rupture and the blood would eventually

boil. However, lower total pressure than atmospheric pressure can keep the astro-

naut alive and be an adequate environment to work in as long as the partial pressure

of oxygen is maintained. One of the most important functions of a spacesuit is to

provide mobility: "the advantage of a human in space over a robot is the ability to

see, touch, and adapt instantly to real-time conditions. This is an advantage only if

the astronauts are able to effectively use their hands, arms, legs, eyes, and brains."

Spacesuit joints are one of the most critical parts of the design of the spacesuit since

they determine its mobility. The common gas-pressurized spacesuit designs tend to

keep a constant or near constant volume in the joints. Spacesuits have evolved since

13

the initial designs but many issues remain. Over time, these gas-pressurized suits

cause fatigue, increase metabolic expenditure, and eventually may lead to injuries

in astronauts. Gas-pressurized suits cause astronauts to experience discomfort, hot

spots, skin irritation, abrasions, contusions, and over time injuries requiring medical

attention [21]. Injuries occur primarily where the person impacts and rubs against

the suit to change its position. Although most injuries have been minor and did not

affect mission success, injury incidence during EVA is much higher than injury that

occurs elsewhere on-orbit. While the most common injuries occur in the hands, feet,

and shoulders, shoulder injuries (including rotator cuff tears that require surgery)

are some of the most serious and debilitating injuries astronauts face as a result of

working in the suit. Countermeasures have been developed to mitigate suit-related

injuries, but still relatively little is known of how humans move within the spacesuit.

In addition to the technical challenges, human spaceflight programs face implementa-

tion challenges in terms of cost, schedule, and regulatory barriers. In order to provide

a holistic view of the problems presented in human spaceflight, we must look at both

the technological and policy aspects. For the purpose of this thesis, we will look

specifically at international cooperation for EVA capabilities.

The objectives of this research are to:

1. gain a greater understanding of the human-spacesuit interaction specifically at

the shoulder interface in order to determine potential for injury by analyzing

the suit-induced pressures against the body using a network of pressure-sensing

and kinematics systems,

2. determine the validity of the particular sensors used in an effort to understand

this interaction and suggest alternatives if the current sensors are found to be

sub-optimal,

3. extend the wearable technology application to other relatable fields such as

soldier armor and protective gear,

4. and finally, compare and contrast EVA capabilities and incompatibilities be-

14

tween the U.S. and Russia in order to build a case for equipment standardiza-

tion.

Chapter 2 provides relevant literature on EVA training, astronaut injury, shoulder

injury, and previous studies and countermeasures addressing these issues.

Chapter 3 sets forth the pressure sensing systems and experimental design that

were used in order to conduct experiments with the David Clark Company for space-

suits and Protect the Force Company for United States Marine Corps (USMC) pro-

tective gear.

Chapter 4 presents the results and discussion of the shoulder pressure data that

was gathered during the human subjects experiment for the spacesuits and protective

gear. A combination of graphical and statistical analyses were performed to examine

the data, and results regarding pressure distributions, pressure profiles, and effect of

pressure magnitudes are presented.

Chapter 5 presents a general comparison of EVA capabilities between the U.S.

and Russia in order to build a case for equipment standardization.

Finally, Chapter 6 provides a summary and conclusion of all the results of the

thesis. Recommendations for future work are also provided.

15

Chapter 2

Literature Review

The following is a review of literature on space suit design and EVA working envi-

ronments.

2.1 Extravehicular Activity

Extravehicular activity (EVA) is critical to human spaceflight. Since the Soviet cos-

monaut Alexey Leonov performed the first EVA on March 18, 1965, to the moment

American astronaut Neil Armstrong stepped foot on the moon on July 20, 1969, the

human race has been eager to further human space exploration for national pride, the

sake of curiosity and possibly survival, as well as secondary rationales that include

economic development, new technologies/innovation, education and inspiration, and

the development of peaceful international relations. Astronauts and cosmonauts have

performed nearly 300 EVAs as of 2009 [6]. Only 14 of those EVAs have been con-

ducted on the lunar surface in one-sixth gravity. However, despite the large number

of EVA expeditions over the course of forty years, relatively little is known about

the interactions between the human and the EVA suit that they utilize. NASA has

already publicly expressed its intentions for human missions to Mars, but before these

missions can be realized, it is imperative that we can thoroughly characterize space

suit performance. This research addresses how we are currently developing a method

to evaluate space suit design and understand injury.

16

Figure 2-1: Extravehicular Mobility Unit and Exploded View Diagram. (ImageSources: NASA. Hamilton Sustrand)

The fundamental challenges faced by U.S. space suit designers include providing

pressure. oxygen, waste removal. coimunication. food, water, therimal control, 111o-

bility. radiation protection, and a safe working environment [20]. U.S. astronauts

currently fly and train in the extravehicular mobility unit (EMU). The EMU consists

of the space suit assembly (SSA), protective and comfort pieces. and the life support

system. The space suit assembly is a 14-layer suit weighing 64 kg (140 lb) that is

pressurized to 29.6 kPa (4.3 psi) [12]. The additional portable life support systeim

(PLSS) backpack increases the total suit weight to 115 kg (254 ib). Components of

the SSA are available in multiple standard sizes to allow astronauts to mix pieces

and provide a better fit. Although suit fit is optimized for each astronaut with the

standard sizes available, not all astroamtsanthropometries cai be comfortably ac-

conmnmodated. Currently, the only component of the SSA that can be custom fit for

an astronatutspecific anthropometric measurements are the gloves. The hard upper

torso (HUT), a fiberglass shell that connects to the arm, helnet. and lower torso

asseniblies. coies in two designs: the pivoted and planar HUT. There are only three

sizes currently available to astronauts: nedimun, large, and extra large [9]. Both are

available for use during training. bitt only the planar HUT is currently used iII space-

flight. The pivoted HUT is no longer used on orbit simnce a rupture in the bellows

would be a catastrophic failure of suit integrity [21]. The planar HUT has planar

17

seve bearings in fixed planes at the armli openings, vhereas the 1)ivoted HUT has a

shoulder gimbal with a two-point pivot to aid the range of motion of the shoulder

joilit [19].

Figure 2-2: Pivoted HUT on left, and Planar HUT on right. Note the different angles

of tlie seve bearings in the two HUTs. (Image Source: NASA)

There are several prototylpe suits that have been developed for planetary aild deep

space exploration. Future nissions to \Jars will require spacesuits that have high

mobility and dexterity. and currentlv the ENJU does not satisfy those requirements.

The E\IU limits mobility and requires a substantial aniount of eiiergy iii order to

move the suit into a desired position. Mechanical counter-pressure suits such as

the BioSiit are currently bei]ig developed to address issues of mobility and energy

consmiiption. but gas-pressurized suits are officially NASA's state of the art capability

[13]. Of the gas pressurized space suits being developed to address different sceliarlo

requirements. the one of particular focus in this thesis is the David Clark Launch and

Entry Development Suit. which is currently being developed to address launch and

entry requirements. No other information is currently publicly available for this suit.

Figure 2-3: David Clark Launch and Entry Development Suit

18

2.2 EVA Training and Injury

Extravehicular activity training currently focuses on preparing only for microgravity

as astronauts are only sent to the International Space Station (ISS). The Neutral

Buoyancy Laboratory (NBL), a 23.5 million liter pool at the NASA Johnson Space

Center, is the primary facility that conducts training for the weightless micrograv-

ity environment; it contains a full-size mock-up of the International Space Station.

Every astronaut spends an average of 11.6 hours of training in the NBL per hour of

planned in-flight EVA and between 200-400 hours in training over the course of their

career [19, 3]. The specific number of hours per astronaut depends on their mission's

specific sorties, and the technical details of the EVA. NASA defined an optimum

work envelope for tasks performed on the Hubble Space Telescope and ISS, however,

repeatedly performing tasks outside the envelope can have a significant impact on the

astronaut [9]. As mentioned previously, the current EMU is pressurized at 4.3 psi,

which makes it difficult to move within the spacesuit. Maintaining different postures

to complete certain tasks within the suit causes increased metabolic expenditure and

fatigue. A combination of the time spent training for each EVA and the EVA suit

mechanics is cited as a contributing factor for the astronaut injury incidence rate

[19, 18]. Gravity acting on the astronaut inside the neutrally buoyant space suit

causes shifting within the suit in the NBL not seen on orbit [3]. Another important

aspect in NBL training is that crews are subjected to inversion for short durations,

where the body is oriented in a head down position greater than 45 degrees as seen in

Figure 2-4 [9]. Inversion is defined as a position in which the body is at a head-down

angle of more than 45 degrees; this position loads the shoulders with the astronaut's

full body weight during training for a few minutes at a time. This commonly leads to

injury and discomfort. The lack of restraint within the suit has resulted in a shoulder

bruising during training for nearly every case. Additionally, insufficient recovery time

between NBL training runs prevents astronauts from physically recovering and may

exacerbate developing injuries [9]. Although no EVA-related injury has prevented

successful completion of a mission objective, there have been several instances when

19

the EVA was nearly terminated(l due to suit discomfort [18 3].

lollA

Figure 2-4:

NASA)Astronaut training in the NBL in an inverted position (hage Source:

Wieii sunt discomfort becomes suit-related injury. it becomes a major cause for

concern. Anecdotal reports from astroiiauts have mentioned the occurreice of in-flight

inisculoskeletal injuries since the begiining of NASA's human spaceflight progran

[18]. In December 2002. Williams and Johnson at NASA created the shoulder injury

tiger team to evaluate the possible relationship between shoulder injuries and EVA

training in the NBL. The Tiger Team confirmed that NBL EVA training was directly

liked to a number of shoulder injuries [9]. By administering aii EMU Shoulder

Injury Survey to 42 astronauts aund astronaut candidates. they were able to find the

primary factors coitributing to both major. defined as significant shoulder injuries

requiring medical intervention or surgical correction. and i inor. defined as self-limited

conditions requiring iiiinimal medical intervention, injuries. They foun11d that factors

contributing to both major andi minor injuries are: limitations to niormal shoulder

mobility iin the E\IU Planar HUT. performing tasks ini inverted body positions. using

heavy tools. and frequent NBL runs. Three astronauts had surgery for EVA training-

related shoulder ilIjuries. only one of which had sustained a shoulder injury previously.

Additionally. the onset of a moderate dull ache over the top of the shoulder or within

20

I

the shoulder joint during or within 24 hours of an NBL run strongly suggests a causal

relationship, and repeated episodes of pain during training suggest overuse that could

lead to surgical repair [21, 9]. The findings and recommendations of the tiger team are

wide-ranging due to the multi-factorial nature of EVA training injuries [21]. Based on

their findings, Williams and Johnson made key recommendations to mitigate injury,

but these recommendations were only based on subjective findings [9].

From 2002 to 2004, Strauss et al. quantified and characterized signs, symptoms,

and injuries resulting from extravehicular activity spacesuit training at NASA's Neu-

tral Buoyancy Laboratory immersion facility. By identifying the frequency and inci-

dence of symptoms by location and mechanism, they determined the most frequent

injuries occurred in the hands and shoulders, with shoulders being rated the most

severe injuries. Of 770 spacesuit symptom questionnaires, 24.6% of tests yielded

symptoms, with 47.6% of symptoms in the hands, 20.7% in the shoulders, and 11.4%

in the feet. The only shoulder countermeasures available are supplementary com-

fort pads, an EMU shoulder harness to prevent shoulder contact complaints and an

optimal suit fit to include unique fit adjustments [19].

In 2007, Scheuring published his results from the Apollo Medical Operations

Project. The Apollo Medical Operations Project collected feedback from 14 of the 22

surviving Apollo astronauts. Recommendations centered on improving the function-

ality of the suit as well as improving human factors and safety features. Of the EVA

Suit recommendations listed, the recommendations related to mitigating astronaut

injury were to improve glove flexibility, dexterity, fit, and to increase general mobil-

ity by a factor of four [17]. The astronauts surveyed also recommended increasing

ambulatory and functional capability through increased suit flexibility, decreased suit

mass, lower center of gravity, and reduced internal pressure [17]. In 2009, Scheuring's

following study cataloged and analyzed all in-flight musculoskeletal injuries occurring

throughout the U.S. space program beginning with the Mercury program through

the conclusion of ISS Expedition 13 in September 2006 [18]. A total of 219 in-flight

musculoskeletal injuries were identified, 198 occurring in men and 21 in women. The

incidence of in-flight musculoskeletal injuries was found to be 0.021 injuries per day

21

for male crewmembers and 0.015 injuries per day for female crewmembers. While

hand injuries represented the most common location of injuries, shoulder and back

injuries are also notable in the data of injuries separated by anatomical location. The

most common types of injury were abrasions, contusions, strains, and lacerations. Of

note, most astronauts also remarked that their wounds healed more slowly while on

orbit. Hand injuries were most common among EVA crewmembers, often due to the

increased force needed to move pressurized, stiff gloves. These hand injuries mani-

fested themselves as small blisters and pain across their metacarpophalangeal (MCP)

joints. Injuries occurred most frequently during crew activity and within the EVA

suit. Engineers can use in-flight injury data to further refine the EVA suit and vehicle

components [18]. In 2010, Opperman developed a musculoskeletal modeling tool to

compare various spacesuit hard upper torso designs and focus on optimizing comfort

and range of the motion of the shoulder joint within the suit. He also performed a

statistical analysis to investigate the correlations between the anthropometrics of the

hand and susceptibility to injury using a database of 192 male crew members' injury

records. He found hand circumference and width of the MCP joint to be significantly

associated with injuries. Experimental testing was also conducted to characterize

skin blood flow and contact pressure inside the glove. The tests show that finger

skin blood flow is significantly altered by contact force/pressure, and that occlusion

is more sensitive when it is applied to the finger pad than the finger tip [16].

Countermeasures to address shoulder injury in the NBL and orbit include different

types of simple padding and harnesses for the HUT. The most commonly used pads

are primarily for protection from the scye bearing and HUT shell, while the rarely

used shoulder harness acts like a pair of suspenders inside the HUT that have a pad

assembly at the shoulders to absorb contact loads in the suit [21]. Countermeasures

can still be improved by expanding our understanding of human-spacesuit interaction.

22

2.3 Previous Work on Development of a Quanti-

tative Understanding of Human-Spacesuit In-

teraction

Over the past few years, a team of researchers at the Massachusetts Institute of Tech-

nology (MIT) along with collaborators at Trotti and Associates, Inc. (Cambridge,

MA), has studied Spacesuit Trauma Countermeasure System for Intravehicular and

Extravehicular Activities under NASA Grant NNX12AC09G. The main objectives

were to: 1) analyze data for correlations between anthropometry, space suit com-

ponents, and injury, 2) model human-spacesuit interaction, 3) design and develop

modular protective devices to mitigate injury, and 4) quantify and evaluate human-

spacesuit interaction using a suite of sensors [14].

The first objective was addressed by a study published in 2014. The study quan-

titatively evaluated the causes of astronaut shoulder injury and performed a meta-

analysis investigating injury trends, proposing an injury classification system, and

creating predictive statistical shoulder injury models using a database of 278 astro-

nauts that included anthropometric measurements, training record, and injury record

[3, 2]. It found that percent of training performed in the planar HUT was the strongest

predictor variable for injury, while training frequency and recovery between sessions

were also important variables. It also identified that bideltoid breadth, expanded

chest depth, and shoulder circumference were the most relevant anthropometric mea-

surements for predicting injury. The second objective was addressed by Diaz at the

IEEE Aerospace Conference in 2014, where a biomechanical analysis using OpenSim

(Stanford, CA) was performed to understand the effect of the space suit on muscle

activation and force generation on the knee using motion capture data and EMU

joint torque data [5]. The third objective was addressed by developing injury pro-

tection concepts, evaluating materials for their offloading capabilities, and eventually

developing both passive and inflatable protective device prototypes. The fourth and

final objective of evaluating human-spacesuit interaction has been addressed, but

23

methods and tools are continually being improved. A human subjects experiment

was performed at David Clark Company and at NASA Johnson Space Center using

the David Clark Mobility Suit and the NASA developmental Mark III suit. A pres-

sure sensing system was built to evaluate pressures over the arm for this experiment,

and the results on sensor performance are analyzed and discussed [3]. An additional

commercially produced pressure sensor measured pressures at the shoulder for this

experiment, and a quantitative analysis of the human-suit interaction at the shoul-

der was published [8]. The two pressure sensing systems were used in conjunction

with kinematic inertial measurement units, and the kinematics data has also been

published [4].

The quantitative techniques used in this thesis and in the preceding research

present a novel way to understand human-spacesuit interaction. Prior to this grant,

studies only included cataloging incidence and mechanisms of injury, but none had

assessed the human-suit interface with experimental methods. The implications of

this research will help to influence suit techniques and future spacesuit design.

24

Chapter 3

Sensor Systems and Experimental

Design

Through prior work on characterizing the human-suit interface, it was determined

that the following suite of pressure-sensing systems and inertial measurement units

were to be used [3, 8].

3.1 Sensor Systems

3.1.1 Low Pressure Sensing System, the "Polipo"

A custom-built, low-pressure sensing system was designed for placement along the arm

for 5-60 kPa pressure ranges [3]. These sensors were created at the Massachusetts

Institute of Technology in conjunction with researchers at the Wyss Institute for

Biologically Inspired Engineering at Harvard. Known as the Polipo, this low pressure-

sensing system uses 12 soft hyper-elastic sensors to measure low-pressures applied to

the body under soft goods. The sensors are cast from a silicon rubber (EcoFlex0030,

Smooth-On, Inc., Easton, PA), and after two individual pieces are mated, they are

injected with a highly conductive liquid metal called galinstan (Gallium-Indium Tin

eutectic, 14364, Alfa Aesar, Ward Hill, MA), in a spiral pattern to minimize strain

readings. Prior to mating the sensors, a flex circuit made of kapton that has been

25

coated and laser cut in a specific circuit pattern is sandwiched between the two

sensors layers. A detailed description of the design and manufacturing process can

be found in reference [3]. These sensors sense pressure through a change in resistance

of the galinstan as the channel walls deflect when normal pressure is applied to the

completed sensor. The change in resistance corresponds to a change in voltage, which

is then calibrated to correspond to the pressure value. Each individual sensor is

housed in a "chele,"which are all connected through the Polipo garment, a wiring

system developed to accommodate system requirements and human range of motion

requirements. The final Polipo design, which integrated seven strands of copper

wrapped polyester per sensor vest, gives the wire a resistance of 0.6 ohms/meter,

while the polyester core allows it to be very durable and flexible, but the wiring itself

was not elastic or electrically isolated. The wire was sewn in a zigzag pattern to

achieve elasticity. The cheles housed the sensors, and the sensors were held in place

by soldering the flex circuit with the copper wiring mentioned above. As the wires

stretch with movement, the ends of the wires are fixed into place with hot glue. The

cheles and the rest of the Polipo were connected to a conformal base layer using

Velcro, and protected by another conformal cover shirt to prevent catching of the

wires or movement of the sensors.

According to Anderson, the pressure-sensing system achieved both high weara-

bility and utility, however, a design concern mentioned for future iterations was im-

proving sensor wiring durability, which proved to be a limitation after several hours

of wear inside the spacesuit performing EVA motions. Tears in the elastomer caused

sensor failure, and the sensors performed sub-optimally under static loading due to

creep effects and hysteresis. Another important area of future work mentioned was

to improve manufacturability such that the process is less highly-skilled, takes less

time, and fewer sensors fail during the construction process.

In an effort to capitalize on Anderson's design suggestions for future iterations of

the low-pressure sensing system, two major possibilities were investigated to address

the durability and manufacturing process concerns mentioned above.

In order to address manufacturing concerns, the possibility of using a 3D-printed

26

sensor. rather thaii a hand-manufactured sensor. was investigated. The W'yss Insti-

tute for Biologically Inspired Engineering at Harvard developed a 3D-printed carbon

grease sen1sor that could efficieitlv and effectively sustain an electric current, shown

in Figure 3-1. Carbon fiber nanotnlbes are suspended in grease. and contact between

the llanotubes creates a complete circuit to sustaim an electric current when a voltage

is applied. After viewing preliminary resistance tests conducted by colleagues at the

WvYss. it was determined that these 3D-printed sensors prvecd to be less reliable ill

their pressure measurements because naliottibe shifts led to inconsistient resistances

across each sensor. This was verified with arbitrary resistance measurements using a

voltmneter. The 3D-printed carbon grease sensors were not pursued.

Figure 3-1: Printed carbon-grease sensor with electrode extensions. (Image Source:

Wyss at Harvard, 2014)

For (irability concerns, a conniercial replacement for the hand-sewn copper

wiring was sewn in order to improve garment elast icitv and iinhnize tearing where the

copper wiring met the sensor. A suitable onie-stranded version of a conductive tape

was found through AMOHR Tecinisehe Textilien GnmbH. a company in Germany

that produces technical narrow fabrics for various purposes. AIMOTAPE Conduct

Nylon + Elastoner #45708. containing 2 insnlated copper strands was custom or-

dered as a replacement for the Polipo. which can be seen iii Figure 3-2. The AIOHR

two-stranded conductive tape was implemented in a new version of the Polipo by

CostnmeWorks in Somerville. MA. The original galinstan sensors, which are difficult

27

to manufacture, were used ilI the second version of the Polipo. however, due to seiisor

nmanufacturing issues leading to sensor failure. the second version of the Polipo was

not tested in the following experiments.

Figure 3-2: AT\(OHR two-stranded conductive tape used for second Polipo iteratioll.

3.1.2 Novel High-Pressure Shoulder Sensor

The Pliance sensing system developed by Novel GmbH. a German comnpaimny that

specializes in dynamic pressure distribution measurement technology, 'an be used

for an accurate mleasiuremlelnt of pressure 1(d load distribution on boti 1(ard and

soft surfaces. The Pliance system was connected to a range of flexible, elastic seil-

sors iade from capacitive transducers with high-tecb elastomers. These sensors are

calibrated through pre-determined loading sequences so as to create a baseline for

future measurements. guarantee accracv amd generate reproduciile data [8]. The

accompanying Pliance software gives the user the ability to acquire and store pres-

sure (istribuition data. view absolute pressure values in each sensor of the sensor ilat

network. playack mileasuremlelnts. and view maxiium1111 pressure. force 81(1 contact

area. The particular sensor used in our experiment and past experiments is a nodi-

fied S2073 sensor mat approximlaitely 22.4 cm x 11.2 cm with 128 individual senlsors

arranged in a grid of 1(6 by 8. Each sensor is 1.4 (111 ill length and width and can

measuire pressures between 20-600 kPa at a resolution of approximlately 1 kPa. The

Pliance systeill uses tell 1.2 V nickel metal hydride batteries with 2000 mAh. and

the sensor is rul at 330 mA. While the data collection rate can be adjusted. for the

purposes of our experiment, the data, was recorded once every 0.02 s (50 Hz). The

28

sensor imat was kept in place using the Polipo's base layer mentioned above, wvhich

was equipped with a rectaigular pocket interface that housed the Novel sensor mat.

Low-Pressure High-Pressure Novel APDNI Inertial

Polipo sensors sensor and hardware 'MeasurementUnit

A) B) C)

Figure 3-3: Experimental Sensor Systemis: A) Low-pressure Polipo sensors, B) High-

pressure Novel shoulder sensor C) APDM Opal inertial measurement unit. (Image

Source: Anderson. 2014)

Prior to any experinients. the Novel sensor is calibrated to ensure accurate data

collection during official mneasurenment trials. The calibration device used ws also

provided by Novel GmbH and was developed specifically for use with sensors devel-

oped by Novel and their Pliance sensing system. The calibration device consists of

an inflatable rubber bladder that is housed by secure rigid plates. The sensor be-

ing calibrated is placed on the calibration board and centered within the alpparatis.

Compressed air is then fed into the device, thereby exerting pressure on the sensor

nat. The Novel software provides caliiration steps to lhad the sensor mat at vain-

ous known pressures in order create calibration curves create( within the software.

Calibration files are stored for subsequent testing.

3.1.3 APDM Inertial Measurment Units

The APDM Opal Inertial Measurement Unit (IMU) Sensing systeii (Portland. OR)

consists of three accelerometers, three gyroscopes, and three nagnetometers. A

Kalnan filter integrates these signals into an orientation quaterion for each IMU.

The IMUs were placed in-phine with one another to optimize the output for isolated

joint movements, but their relative orienitations allow the detection of off-axis rota-

tions [3]. Three sensors were mounted internally on the upper arm. lower arm. and

29

Low- Pressure Polipo --Sensor Network

High-pressure NovelSensor Mat

Body MountedOpal IMUs

Figure 3-4: Placeineit of the ill-suit sensor systems. (Image Source: Anderson. 2014)

chest. Three cxternally imounted sensors were correspondingly iouinted oil the up-

per and lower spacesuit ami and suit torso. Each sensor is 4.8 x 3.6 x 1.3 ciii and

weighs approxiinately 21 g. Tl he gyroscopes and imagnetonieters were recalibrated

before placed on each sul1ject to take into account the iagnetic environment and

inimilize the gyroscope drift over tiie. They are powered by a lithiuiI battery at 3.7

V nloiliniaI, nd the imiaximumiliii current through the sensor is approxiimately 56 mA.

IMU seisoi data was collected wirelessly and continuously synchronized in real time.

3.2 Spacesuit Testing Experimental Design

This experiment was performed using two subjects in the David Clark Launch and

Entry Development Suit. The suit was pressurized aid tested at venting pressimre

(0.25 psi). iiltermedliate pressure (2.5 psi). and full pressure (:3.5 psi). While the

EIU defines 4.3 psi as "full pressure . David Clark pressurizes their suits to 3.5 psi

ill order to iicrease mobility 1)it ,minitai a smaller safety mgin for oxygen partial

pressure requirements. They wvere asked to perforn a series of upper body notions

inside the spaceslit while lying iii the recumbent position. These series of upper-

body motions is niilmed at characterizing') the hunan-suit interactions. Three isolated

joint nmovenmenits vere evaluated: elbow flexion /extension. shoulder flexion/extension.

and shoulder abduction/adduction. i addition. one nmulti-joint functional task was

evalulated: the cross-body reach.

Elbow Flexil/Extenslo VThe subject stands away from the donning stand supported by theirown effort. Beginning with both arms relaxed at their side, palms k s Dofacing anterior, the subject bends the anis at the elbow through s 'their maximum range of motion. The subject then releases to therelaxed position. MO M

Shoulder Fexion/LxtensionThe subject stands away from the donuing stand supported by theirown effort. Beginning wsith both anns relaxed at their side, thesubject bends the arms at the shoulder through the sagittal plane.The subjects move through their maximun range of motion. Thesubject then releases to the relaxed position.

Shioulder Abduction/AdductionThe subject stands away from the donning stand supported by theirown effort. Beginning with both arns relaxed at their side, thesubject bends the anns at the shoulder through the coronal planeThe subject moves through his or her mAximumi rangP ot motionThe subject then releases to the relaxed position

Cross-Body ReachThe subject begins in a relaxed position and reaches across theirbody to touch their hip on the opposite side. The subject mos atheir ann up to chest level and sweeps in front of their body. Whenthe arm is extended in front of the shoulder, the subject touches thehelmet on the same side The niovement is then repeated with theopposite arm.

Figure 3-5: Descriptions of the four upper body motions performed during the space-

suit experiment: three isolated joint notions (elbow flexion/extelision. shoulder flex-ion/extension. shoulder abduction/adduction) and one functional task (cross bodyreach). (Image Source: Anderson. 2014, Hilbert et al. 2014)

The test protocol consisted of 15 repetitions of the four different motions inside

the spacesuit. These repetitions were divided into three groups of five repetitions

to allow for assessment of fatigue or changes ill biomnechanical strategies.

XVere divided into movement groups such that the order was counterbalanced within

the groumlp [3]. Prior to the test. subjects were trained on each motion and allowed

to practice it until they were comfortable ill order to maximized mnotion coilsistency

duirilng the experiment. The subject performned each notion iii the prescribed order

of the movement group, with no less than a 5-minute break between each movement

group in order to (ollect subjective feedback and to allow the subject to rest. After

all three imovment groups were completed, there was an intermittent rest period

to increase the pressure iii the suit. The subject was first tested in the unsuited

condition, and then at the corresponding test pressulr('s in the suit. The pressure

profiles and joint angles were recorded throughout the experiment. A representative

experiment schemlati( is showii in Figure 3-6, and the full experimental test plan can

31

I

Motionls

Movement Group 1 Movement Group 2 Movement Group 3

H 11 12 13 14 13 I1 14 12 12 14 11 I3

11: Elbow Flexion/Extension12: Shoulder Flexion/Extension Si vvvr.13: Shoulder Abduction/Adduction14: Cross Body Reach S2 r

Figure 3-6: Experimental Design Test Protocol: Each imovenent group consists of a

counterbalanced ordering of the motions. The motions studied were: three isolated

joint iot ions (elbow flexion/extension. shoulder flexion/extension. and shoulder ab-

duction/adduction) and a fiictional task motion (cross body reach). In each move-

iment group. the specific nmotion was repeated 5 tinmes for a total of 15 repetitions per

mlotionl.

3.3 Marine Protective Gear Experimental Design

In an effort to expand the applicalbility of the sensor systems. two rounds of experi-

ments were performed in conjunction withIi Protect the Force. a strategic consulting

firm specializing in product development for the U.S. aried forces.

Infantry soldiers and officers are a central comnponent of ground forces in the Ma-

rine Corps anod other branches of the military. According to the Marines, infantrymen

are trained to locate, close with and destroy the enemyn 1y fire and umaneuver, or repel

the eiemv's assault by fire aiid close co1bat. Riflenem serve as the primary scouts.,

assault troops and close colmlbat forces within each infantry unit. Crucial to combat

iission effectiveiness is ensuring each Marine's safety. However. in order to provide

safety in the form of heavy armiiior, often physical mobility aid strength must be less-

ened or compromised to carry heavy loads of arimor in addition to the gear Marines

are required to carry. Iii an effort to provide lightweight but effective armmor to lesseii

32

he found inl Appendix A.

heavy loads and increase imobility while wearing armor. the Marine Corps System

Command has developed several prototypes of advanced larine protection gear as

alternatives to the current gear provided to Marines.

Both experiments were performimed with the same subject ill (lifferelt proteetion

gear configiiratioiis: 1) the interim capability, USMC Plate Carrier (PC) and neck

plates. 2) the current capability. USNIC Improved Modular Tactical Vest (IMTV), 3)

the newly designed Ballistic Base Layer (BBL) protective garment and 4) the future

capability, the plate carrier combined with the BBL protective garment. Different

configurations of the protection gear were tested for their mobility. the shoulder con-

tact pressure, aid the subjective evaluation for comfort. fatigue, and mobility. This

is critical in order to ensure the future capabilities being currenItly developed will

provide an improvement in the design and the use of the protection gear.

Figure 3-7: Different US\IC protection gear conifiguratiois used during testing

One subject was tested in the Man Vehicle Laboratory. at the Massachusetts Iii-

stitute of Technology (Cambridge, MA) oil two different occasions (December 2015

and Jaimuary 2016). The subject corresponded with the Narine infantrymen anthro-

polmetrics provided by Protect the Force as seen iln Table 3.1.

Similar to tlie spacesulit test experimental desigm. the experiment colmsisted of

15 repetitions divided into three groups of five repetitions of four different mo-

tions inside the spacesuit. Three isolated joint movements were evaluated: shol-

33

Table 3.1: Anthropometrics from typical Marine compared to anthropometrics of

subject usedMarines [ Subject

Height 5'8" 5'11"Weight 176 lbs 190 lbs

Waist Circumference 34.5" 34.5"Chest Circumference 40.5" 38"

der flexion/extension, the vertical shoulder abduction/adduction (defined as simply

the shoulder abduction/adduction in Figure 3-5, and the addition of the horizon-

tal shoulder abduction/adduction shown in Figure fig12. The multi-joint functional

cross-body reach was also evaluated.

A

SHOULDER ADOUCTION (A),ABDUCTION (B)

Figure 3-8: Horizontal shoulder abduction/adduction. Beginning with their arms

extended at shoulder height, shoulder width apart, the subject bends their arms at

the shoulder in the transverse plane. The subject moves through his or her maximum

range of motion. The subject returns to the initial start position then releases to the

relaxed position.

This motion is particularly useful to combine basic motions performed by Marines

during operations: reaching helmet on its side, reaching opposite side of the body, or

extending the arm in front of the body. The subject performed each motion in the

34

prescribei order (Iof the movement group. with n1o less than a 5-iiiiinute break 1b)etweeli

each movenient group ill order to collect subjective feedback an( to allow the subject

to rest. After all three inoveiiient groups were conmlleted(. there was ain iiiteriiiittenit

rest period. The subject was first tested( in the unsuited condition. ald then with the

corresponiiig protective gear coiifiguratioiis. At the en(l of the second experinelit

following all iotiolis outlined in the experimienital design, the subject )erufoled a

set of rifle carry motions in Configurations 1 an(l 4. The rifle carry iotions were

perforimled ill the followillg sequence for each condition: 5 repetitions. 10 second pause,

aild 5 repetitioils for a total of 10 rifle carry repetitions per configuration. The rifle

(ally motiol1s were performed with a 7.2 lb pipe similar in shape to an M16 A2 rifie.

The shoulder contact pressure profiles al(l joint angles were recorle( trlonghoIt the

experimeiit. A representative experiment scheiatie is shown in Figure 3-9.

Movement Group 1 Movement Group 2 Movement Group 3

11 12 13 14 j3 11 14 12 12 14 11 13

11: Shoulder Flexion/Extension12: Shoulder Abduction/Adduction vertical S1 / \/\A13: Shoulder Abduction/Adduction horizontal14: Cross Body Reach S2 V0

Figure 3-9: Protective Gear Experimental Design Test Protocol. Each movement

group consists of a couiiterbalanced ordering of the motions. The motions were: three

isolated joint motions (shoulder flexion/extension, shoulder abduction/adductioli ver-

tical. shoulder abduction/adduction horizontal) and a functional task motion (cross

body reach). In each imoveineit group. the specific motion was repeated 5 times for

a total of 15 repetitions per motion.

35

Chapter 4

Novel System Results and

Discussion

The following chapter presents a variety of analysis intended to provide an under-

standing of the human-suit shoulder interface. The diagram in Figure 4-1 shows the

presentation of the data and the orientation of the Novel sensor with respect to the

subjects shoulders. This diagram shows that the lower portion of the sensor with

respect to the diagram corresponds to the anthropometric region toward the clavicle

and front of the body, whereas the upper portion overlays the back of the shoulder,

toward the shoulder blade.

4.1 David Clark Experiment

For the David Clark Launch and Entry Development Suit, two main categories of

data are presented for the Novel pressure sensing system: 1) the overall pressure dis-

tributions and 2) the pressure profiles seen in each of the motions. For the following

results, data from the elbow flexion/extension was excluded as it was deemed less

relevant to the shoulder portion of the human-suit interface. Hilbert first determined

that analysis of pressure distributions aids in determining which areas of the shoulder

are experiencing the highest pressures during upper body motions and providing a

visual understanding of what is happening at the human-suit shoulder interface. Ad-

36

Diktal fnd of Shoulder Ba dg i u w suusc 111m Im I

Figure 4-1: Orientation of the Novel Sensor. The orientation Corresponds to the

information i i each pressure (listtribution figure. Coloring is for orientation and is not

related to pressure scales. (Image Source: Hilbert 2015)

ditionally, analyzing pressure profiles as a function of tie provides how the pressures

vary over the course of a particular movement while allowing us to determine whether

there is any time effect [8]. Upon visual inspection we can claim that Subject 1 had a

narrower an(d taller body frame than Subject 2. which will be critical to the discussioli

of varying contact pressures.

4.1.1 Pressure Distributions

The pressure distribution mnaps for Subject 1 and Subject 2 are shown in Figures 4-2

and 4-3. The figures show the pressure distributions as a color scale representing the

pressure in kPa. For practical reference. 100 kPa provides approxinately the same

pressure as a 1 kg (2.2 lbs) weight oi one square centimeter of the skin. For each

of the motions at each of the pressurized conditions. the pressure (listributioli map

represents the pressure (listribution at the peak of the movenient, or the pressure

distribut ion at the moment when the highest pressure appeared.

Looking at Subject I's neasuireients in Figure 4-2. it is evident that as suit

presslurizatloll increases. contact pressure increases as well. From visual inspection,

it is not clear whether any one motion has higher Overall pressures than the other

motions. At 0.25-psi vented condition. pressure is concentrated along a line just above

37

Subject 1

Sh FI/Ext Sh Abd/Add CrBReach

IA

Ln~

4,)

CL

(n

PI

I

I

160140

120100

80

6040

20

0 kPa

160140

120

10080

6040

200 kPa

160

140120

100

8060

40

20

0 kPa

Figure 4-2: Contact Pressure listributions for each inotion at ( (ch Cressurized eon-dition for Subject 1.

38

the clavicle for all three motions, likely over soft musculature near the top of the

shoulder as was seen in Hilbert's results. These areas of pressure concentration (peak

of ~75 kPa) are accompanied by a secondary area of pressure concentration at the

most distal end of the lower edge of the sensor. Likely, the sensor was being slightly

pinched by the chest and the armpit in each of the motions. However, after dynamic

inspection of the data in the form of pressure distribution videos, the pressure line can

also be attributed to a crease in the Novel pressure sensor at the peak mobility point

of a motion. During motions where the subject retains high mobility, a crease would

form in the mat at the top of the shoulder due to the rigid nature of the mat, and

this produces artificially high pressure values. At the 2.5-psi intermediate-pressurized

condition, the pressure (peaks between -100 and ~140 kPa) is now concentrated in a

region centered just above the end of the clavicle toward the acromion for all motions.

At this pressure and higher pressures, the Novel mat did not seem to demonstrate

any creasing, most likely due to limited mobility and the mat pressing against the

suit. The peak pressure is highest for the cross body reach, followed by the shoulder

abduction/adduction, and lastly the shoulder flexion/extension, but the location and

shapes of the pressure distributions are nearly identical in all motions. At the 3.5-psi

full-pressurized condition, the pressure distributions maintained the size and shape

of the peak pressure locations for the intermediate pressure condition, however, the

magnitudes of the peak pressures reach -160 kPa at the clavicle and acromion.

Looking at Subject 2's measurements in Figure 4-3, the pressure distributions

follow similar trends to Subject 1's such that as suit pressurization increases, con-

tact pressure also increases. At the 0.25-psi vented condition, the artificially high

crease pressure band is only seen in the shoulder flexion/extension. For the shoulder

abduction/adduction and cross body reach, there is a large but low-pressure (-40

kPa) peak at the top of the shoulder toward the clavicle and chest. At the 2.5-psi

pressurized condition, the size of the peak pressure area is reduced and becomes

more concentrated in the acromial region toward the distal end of the shoulder for

all motions. The shoulder abduction/adduction experiences the highest peak pres-

sure (-120 kPa), followed by the shoulder/flexion extension (-100 kPa), and the

39

Subject 2

Sh Fl/Ext Sh Abd/Add CrBReach

En

Lfl

CL

U')

C0

I

I

I

140

120

100

80

60

40

20

0 kPa

140

120

100

80

60

40

20

0 kPa

140

120

100

80

60

40

20

0 kPa

F'igiire 4-3: Contact pressure distributions for each 1i1otion at each pressurized cou-

dition for Subject 2.

40

cross body reach (~70 kPa). For the 3.5-psi pressurized condition, the size, shape,

and location of the peak pressures remains almost identical to the 2.5-psi pressurized

condition. The peak pressure locations are concentrated at the top of the shoulder

toward the acromion for all three motions. The peak pressure magnitudes for the

shoulder abduction/adduction and cross body reach approximate to ~120 kPa, with

the shoulder flexion/extension peak pressure magnitude reaching -140 kPa at the

acromion.

Comparing subjects we see that Subject 1 experiences higher pressure than Sub-

ject 2 in all motions at the fully pressurized condition (3.5-psi pressurization). It

is interesting to note that for both subjects, pressure was concentrated in a consis-

tent location across all motions: approximately on the acromion and just above the

clavicle and the soft musculature at the top of the shoulder.

4.1.2 Pressure Profiles

Pressure profiles as a function of time are now considered at each pressurized condition

for each subject, shown in Figures 4-4 through 4-7. For each motion, selected pressure

profiles for different sensors are plotted for each of the three movement groups. The

individual sensiles are chosen based on whether they experienced the peak pressure

on the mat at any moment in time during the motion repetition, and they represent

the highest magnitude profiles of each general trend of sensor response. Since it

was determined that at lower pressurized conditions where the subject retains high

mobility the Novel mat develops a crease and reads artificial pressures, the pressure

profiles of the lowest pressurized condition, 0.25-psi, are not shown.

All plots have the same scales: the y-axis being pressure in kPa from 0 to 160 kPa,

and the x-axis being a normalized time axis. Each cube in the grid represents a 0.1

sec interval in the horizontal direction and a 20-kPa interval in the vertical direction.

Normalizing the x-axis and plotting each of the profiles on the same time scale allows

for easier comparison. While each motion included five repetitions per movement

group, only the two most consistent repetitions are shown since the subjects found

it very difficult to remain consistent in the recumbent position. All motion pressure

41

U'

Subject 1- 2.5 psi pressurized conditionMovmt. Group 1 Movmt. Group 2 Movmt. Group 3

-&j

.0

U

U

.1

11I-t

A- -&jr NFigure 4-4: Contact pressure profiles for all motions in the 2.5-psi pressurized condi-tion for Subject 1.

profiles are shown, however in some cases, it is impossible to identify the profile of

the motion.

Starting with Subject 1, we will analyze the pressure profiles by motion. The

shapes of the general profile for the shoulder flexion/extension are consistent for both

the 2.5 psi pressurized condition and the 3.5 psi pressurized condition. The shoulder

flexion/extension appears to have two distinct peaks per repetition in approximately

the same location at the top of the shoulder where the second peak is only a sensile

or two closer to the chest than the shoulder blade. Analyzing the subject video

taken during the experiment, it appears that the subject would shift inside the suit

during the beginning of the motion during flexion, and then shift again after the

peak of the motion during extension, hence the slightly higher contact pressures

seen in the second peak. The shift in full body position can be attributed to air

displacement in the soft suit in the recumbent position since the subject did not have

42

El

Subject 1- 3.5 psi pressurized conditionMovmt. Group 1 Movmt. Group 2 Movmt. Group 3

UJ

CA

*I t

Vt t tj~J4M

ra

L) 4!k

Figure 4-5: Contact pressure profiles for all motions in the 3.5-psi pressurized condi-tion for Subject 1.

43

-I

the stability of standing on his feet. There is also a constant contact pressure seen

at the top right corner, which corresponds to the shoulder blade. This is due to

resting on the shoulder blade and back in the recumbent position and the shoulder

activity that occurs in the shoulder blade during the shoulder flexion/extension. For

the shoulder abduction/adduction, the general profile for the peaks are consistent,

however, the magnitudes of the peaks in the 3.5 psi pressurized condition are highly

inconsistent between movement groups. During the shoulder abduction/adduction,

the same body shift due to air displacement in the suit occurred as it did in the

shoulder flexion/extension. There is an initial spike in contact pressure as the body

shifts toward the feet in the suit and initiates contact at the top of the shoulder, then

again as the body shifts back upward toward the head and initiates contact during

the contrary movement. In the cross body reach movement, the pressure profiles

change in between pressurized conditions. In the 2.5 psi pressurized condition, there

are three peaks: the two larger peaks occur at the top of the shoulder during the

motion, and there is a much smaller peak between the two larger peaks that occurs

at the shoulder blade. These peaks coincide with the multiple motions necessary to

complete the functional task motion. In the 3.5 psi pressurized condition, the same

two major peaks occur without the smaller peak occurring at the back of the shoulder.

Looking next at Subject 2's pressure profiles, the shapes of the general profile for

the shoulder flexion/extension are consistent for both the 2.5 psi pressurized condition

and the 3.5 psi pressurized condition. Unlike Subject 1, the shoulder flexion/extension

appears to have only one distinct peak per repetition in approximately the same lo-

cation at the top of the shoulder as Subject 1. While the same body shift was seen

inside the suit as Subject 1, due to the different anthropometries between subjects,

the suit displacement had less of an effect on Subject 2 since Subject 2 had over-

all larger anthropometric measurements. For the shoulder abduction/adduction, the

general profile for the peaks are consistent and nearly identical to the shoulder flex-

ion/extension profile. Before the large prominent peak, there is a smaller, less distinct

peak of contact pressure that occurs on the back of the shoulder toward the armpit

(top left corner of the diagram) during the 2.5 psi pressurized condition. This small

44

Subject 2- 2.5 psi pressurized condition

Movmt. Group 1 Movmt. Group 2

B~11wi

Movmt. Group 3

. rf.

Figure 4-6: Contact pressure profiles for all motions in the 2.5-psi pressurized condi-

tion for Subject 2.

back-of-shoulder peak also occurs during the 3.5 psi pressurized condition, but not in

all three movement groups. In the cross body reach movement, the pressure profiles

remain extremely consistent between pressurized conditions, only the magnitudes of

the first large peak changes. Subject 2's cross body reach experiences only two ma-

jor peaks instead of three as seen in Subject 1: initial large park at the top of the

shoulder, and a second smaller peak occurring at the back of the shoulder blade next

to the arm pit.

Comparing the pressure profiles between subjects, it appears that while the general

pressure distributions appear to be similar, the pressure profiles give the resolution

to observe distinct differences between subjects. The primary location for contact

pressure in Subject 2 is located slightly to the left of the primary location for contact

pressure in Subject 1. This can be for one of two reasons: either the mat placement

was slightly different between subjects and so the primary locations for both subjects

45

XLLI

.0

U

Movmt. Group 1

Subject 2- 3.5 psi pressurized conditionMovmt. Group 2 Movmt. Group 3

-J

t

4*1

r rNZ23A

Figure 4-7: Contact pressure prtion for Subject 2.

ofiles for all motions in the 3.5-psi pressurized condi-

6

xLU

-

-o

at

U fa It

corresponds to the same anthropometric location on the body (the acromion), or the

differences in anthropometric measurements between subjects caused the primary

contact pressure location to vary slightly between subjects but remain in the general

area at the top of the shoulder. The only motion profiles that vary significantly

between subjects are the pressure profiles of the cross body reach motions between

subjects. While the two peaks seen for Subject 1 are both concentrated at the top of

the shoulder, the two peaks for Subject 2 shift from the top of the shoulder to the back

of the shoulder blade. This supports the claim that for functional movements, contact

pressure locations can vary depending on anthropometric measurements. Another

example is demonstrated during the instances in which there is contact with the

shoulder blade: Subject 1 tended to experience contact pressure on the inside of the

shoulder blade toward the spine, whereas Subject 2 experienced contact pressure on

the outside of the shoulder blade toward the armpit.

4.1.3 Statistical Analysis

In order to more clearly understand the effects of pressurization conditions, motion

type performed, and subject variability, a statistical analysis was performed. Five

peak pressure values were extracted from the data for each motion during each move-

ment group. Each subject had a total of fifteen peak pressure values for each motion

during each condition. The mean and standard deviation were calculated for the peak

contact pressures, and a statistical analysis was performed. A multi-factor ANOVA

(Factor A- motion, Factor B- pressurization condition, Factor C- subject) was per-

formed, as well as Kruskal Wallis tests since shoulder abduction/adduction at 0.25

psi and cross body reach at 2.5 psi for Subject 2 were not normally distributed. For

all tests, an alpha value of 0.05 was used to determine significance.

First, we will analyze the effect of pressurization on mean contact pressure. Main

effects for pressurization condition (p

shoulder flexion/extension, both subjects experienced a significant change in contact

pressure between vent pressurization and full pressurization as well as intermediate

pressurization and full pressurization. For the shoulder abduction/adduction, Subject

1 experienced a significant change in contact pressure between all three conditions,

whereas Subject 2 did not experience any significant change in contact pressure be-

tween the intermediate and full pressurization conditions. Finally during the cross

body reach, Subject 1 experienced a significant change in contact pressure between

vent pressurization and full pressurization as well as vent pressurization and inter-

mediate pressurization, but no significant change was found between intermediate

and full pressurization. Subject 2 experienced significant changes in contact pres-

sure across all conditions during the cross body reach. Since the pressure profiles

reflect the same data as the pressure distributions seen earlier, all vent pressure pro-

files recorded and considered in the statistical analysis are likely higher than actual

pressure experienced by subjects since peak data reflects pressure during the mat

crease.

Next, we will analyze the effect of motion type on mean contact pressure. Main

effects for type of motion were not found (p=0.73) for the multi-factor ANOVA.

The results are presented in Figures 4-10 and 4-11. There is also no single motion

that produces higher contact pressure than any other motion across all cases. At

vent pressurization, Subject l's shoulder abduction/adduction contact pressures are

significantly lower than the other two motions. However, at the same pressuriza-

tion, Subject 2 experiences the highest contact pressures during the shoulder flex-

ion/extension and the lowest contact pressures during the cross body reach, those

being significantly different from each other but neither from the contact pressures

found during shoulder abduction/adduction. It is difficult to determine the validity

of motion type results during the vent pressurization condition due to artificial mat

pressures, so while there are no conclusions to be drawn from the results, the re-

sults would not be considered for recommendations. At intermediate pressurization,

there are no significant differences in contact pressure between any of the motion

types. At full pressurization, the shoulder abduction/adduction differs significantly

48

2

0101

13

0 Julder Flex)ExtL r'lrAbdSAdd

7 rssbody Reach

60 -

20

00-

40

20-

Vent Pressure Intermediate Pressure

Figure 4-8: Effects of pressurization on imean contact pressure for Subject 1.

J43-

Shoulder Flex-ExtShoulder Abdc i

---iCrossbody' Reach

100

2 IsO

4u

velit Pressure Irterr'rede Pressure

Figure 4-9: Effects of pressIIrizatioII OH mIea coutact l)ressure for Subject 2.

49

Subject 1

Full Pressure

Subject 2

Full Pressure

250t Subject 1Vent Pressure

viIntermediate PressureFull Pressure

200C

1501

0

0Rud Fle Ex r ouidr AtxdAdd r' dy Reach

Figure 4-10: Effects of motion type on mean contact pressure for Snbject 1.

Subject 2

Vest PressureI1Intermediate Pressure-Full Pressure

IL

Shoulder FlexiExt Shoulder AbdAdd Crossbody Reach

Figure 4-11: Effects of motion type on mean contact pressure for Subject 2.

50

1 2

intermediate Pressurization

60

40

Shoulder Flex,.Ext Shoulder Abd/'Ad Crossoody Reach

Figure 1-12: Effect at subject on mean contact pressure for iflterllediate pressuliza-tion.

Full Pressurization

- ubI IDuI 2

200 -

IOU -

Shouldei Flex Ext Shoulder Abd/Add Crossoody Reach

Figure 4-13: Effect of subject on iean contact pressure for full pressurizationl.

11

51

d

from both the shoulder flexion/extension for both subjects. However, in Subject l's

case it is significantly higher than the other two motion types whereas Subject 2's

case demonstrates that it is significantly lower than the other two motion types. It

can be said that at intermediate pressure the contact pressures experienced are all

similar between motions types, and at full pressure the contact pressures experienced

are similar between the shoulder flexion/extension and the cross body reach but not

the shoulder abduction/adduction.

Finally, we will analyze the effect of subject variability on mean contact pressure.

Main effects for different subjects were found (p

4.2 Protect the Force Armor Gear Prototype Ex-

periments

The advanced Marine gear aims at distributing loads and pressures more evenly across

the shoulders as opposed to having concentrated areas of extreme pressure at the top

of the shoulders. The Novel pressure sensor was located at the top of the shoulder,

and the data will be displayed in an identical fashion to the shoulder data from the

David Clark spacesuit shoulder experiment.

The pressure distribution maps for the different configurations are shown in Fig-

ures 4-14 and 4-15. The figures show the pressure distributions as a color scale

representing the pressure in kPa. For each of the motions at each of the pressur-

ized conditions, the pressure distribution map represents the pressure distribution

at the peak of the movement, or the pressure distribution at the moment when the

highest pressure appeared. The movements that provide the most insight on changes

in pressure distributions across the shoulder are the vertical and horizontal shoulder

abduction/adduction movements.

Figure 4-14 shows the pressure profiles during moments of peak pressure for the

four separate suited configurations and four separate motions in kPa. The top two

configurations are the configurations without the BBL (the potential future capabil-

ity). When the Novel mat is used on top of the shoulder during motions of high

mobility, it causes the mat to bend, which is shown across all four conditions as a

diagonal increase in pressure across the mat. Configuration 2 (frog -shirt + IMTV)

show the highest overall pressures distributed across the top of the shoulder. The

more lightweight current capability, Configuration 1 (frog shirt + PC), also shows

heavy pressures across the top of shoulder as well as mat bending. The most signif-

icant comparison to make is between Configuration 1 and Configuration 4; both use

the PC but the BBL has also been incorporated into Configuration 4. The pressure

distribution across the shoulder is similar, however, the pressures in Configuration 4

are much lower overall across all sensors. This indicates the BBL has relieved the

wearer of some of the pressure/weight from the PC.

53

onCCt}

onICCC

(-7

rnor

ICCCU

on42C0

(-7

Sh. FI/Ext.

It'I I-T!

ShAbd/Add V. ShAbd/Add H. CrB Reach

fl pi~i~..m

J-t 71W-i'1-

A j.4. I

I

U

1009080

70

60

5040

3020

100 kPa

100

9080

7060

5040

3020

100 kPa

100

90

80

1060

5040

3020

100 kPa

100

908070

60

5010

3020

100 kPa

Figure 4-14: Pressure distributions of all motions for different US\JC arnior configu-

rati(1s.

During the horizontal aidiction/adhuetion. the weight of the arumor is plaeed

heavily (1 the shoulders. aid depell(ling on the widtll of the straps. either manifests

itself as an acute pressure point or more even distribution. In Configuration 1 anld

Configuration 2. the pressure distribution shows the iiat bendiig phenoiiienon. Ili

addition. the surrounding pressures are higher with the frog shirt and PC thian the frog

shirt and IMTV. While the IMTV is heavier., it is most likelv that the wider straps

of the INITV cause a wider (listributioli of pressure of the armor weight, thereby

relieving the subject of a concentrated pressure.

10090

8070

g 6050403020100 kPa

Figure 4-15: Pressure distributions for rifle carry motion for Configuration 1 (on left)

and Configuration 4 (on right).

Figure 4-15 shows the pressure distributions at the mnoment of peak pressure for ri-

fle movements between the two different PC configurations. The figure clearly delon-

strates that the BBL helps minimize pressure at the top of the shoulder when usiig

the PC. The frog shirt loes little to nmninze pressiures at the tops of the mnotionis.

Overall, it appears that the IMTV distriblites the weight of itself better than the PC

oii its owii does, eveni though the PC is 1mch lighter. However. the addition of the

BBL reduces the load of the PC ini some cases and is not helpful in others.

4.3 Conclusions and Future Work

These results yield no common conclusions across suit types or subjects. For the

David Clark Lauinch and Entry Development Suit. both subjects experieiice(d sig-

nificant contact pressures at the top of the shoulder and acromion. However, the

magnituides of contact pressures were significantly different between the two suilbjects

and furthermore, it was not clear whether certain imotions elicited more contact than

others. The effects of motion type oi contact pressure cannot be generalized across

subjects as they are likely affected by individual anthropomnetry, suit fit. and bionie-

chanics, but the information gathered for each subject can be used to decrease the risk

of astronaut injury when applied individually. The less experienced subject (Subject

1) experienced the highest pressures, but both subjects experienced discomfort on

the top of the shoulder over time.

The results yielded for the Protect the Force armored gear can draw some general

conclusions, but since only one subject was tested, further testing is necessary in order

to validate these conclusions. While heavier, the IMTV provides a better pressure

distribution than the PC due to its wider straps. When the BBL is incorporated,

it has the same effect as the IMTV (in terms of distribution) by distributing the

PC load across the shoulder. It also appears that this load exerts an overall lesser

force than the IMTV. The vertical abduction/adduction causes pressure across the

shoulder between all motions regardless of configuration. The IMTV and frog shirt

seem to provide an overall less pressure than the PC and frog shirt. During rifle carry

motions, the BBL significantly offloads pressure from the PC as compared to the frog

shirt.

The most important point to address is the validity of the results. While the

Novel sensor system proves to be state-of-the-art pressure sensing equipment, it may

not be the optimal equipment for the particular human shoulder application. When

placed at an interface with high mobility, the sensor is susceptible to false, artificial

readings caused by creases in the sensor. As a future alternative, it would be ideal

to incorporate a network of small, variably placed sensors across the joint and rest of

the body. A sensor such as the Novel S2012 shown in Figure 4-16, which is 2 cm in

diameter, if paired with many others, could get a general profile for pressure readings

across the shoulder while remaining small enough to gather data across a seemingly

flat surface.

Further studies should integrate the spacesuit pressure and joint angle data found

in other work with metabolic data in order to understand how fatigue and injury in-

fluence the metabolic work necessary for spacesuit operations. All of this information

would allow us to more accurately determine where injury is most probable, incorpo-

rate a quantitative measurement for fatigue, and ultimately influence air-pressurized

56

spacesuit design in the future.

Figure 4-16: Novel Single S2012 Sensor with 2 cin diaieter (Image Source: iiovel.(e)

57

Cha

Signature redacted-- Siqnature redacted

Documents

_Signature redacted Signature redacted

acted Signature redacted

;) Signature redacted

Scenic Quality of ManjéJorqsts Signature redacted for...

Signature redacted redacted I

Author... Signature redacted Sianature redacted

Signature redacted redacted

Signature redacted Professor of Political Scienc and...

Signature redacted - DSpace@MIT

. redacted *' redacted

Signature redacted for privacy. Dr. Ira W. Deep

Signature redacted' Thesis - Massachusetts Institute of ...