ABSTRACT 
 
Title of Proposal: THE EFFECT OF INSPIRATORY AIR HUMIDITY 
AND TEMPERATURE ON PERFORMANCE TIME 
WHILE WEARING A RESPIRATOR 
 
Master of Science, Erica B. Francis  
 
Directed By: Biological Resources Engineering 
Dr. Arthur Johnson 
 
Respirators are worn in only about 20-30% of the appropriate circumstances.  
This research examined the effect of inspired air conditions on performance time.  An 
environmental chamber supplied air at 27?C, 28?C, 37?C, and 55?C to a powered air-
purifying respirator (PAPR) worn by the subject while exercising in a neutral 
environment.   A mathematical model of performance time as a function of the heat index 
(HI), respirator familiarity (RF), personality type (SN), and minute volume (V
e
) indicated 
that performance time increased with a decrease in HI and V
e
and with an increase in RF 
and SN ratio.  A model of performance time as function of user acceptability and the heat 
index indicated that time to reach a level of ?fairly uncomfortable? decreased 
exponentially from 5.34 minutes to 2.85 minutes with an increase in the heat index.  
Performance at the heat index conditions may be described by physiological, subjective, 
and individual characteristics. 
THE EFFECT OF INSPIRATORY AIR HUMIDITY AND TEMPERATURE 
ON PERFORMANCE TIME WHILE WEARING A RESPIRATOR 
By 
 
Erica Bicksler Francis 
 
Thesis submitted to the Faculty of the Graduate School of the  
University of Maryland, College Park, in partial fulfillment 
of the requirements for the degree of 
Master of Science 
2006 
 
Advisory Committee: 
Professor Dr. Arthur Johnson, Chair 
Professor Dr. Gary Felton 
Professor Dr. Fredrick Wheaton 
 
? Copyright by 
Erica Bicksler Francis 
2006 
 
ii
Dedication 
 
To JR and Grace 
 
iii
Acknowledgements 
 
I would like to thank Dr. Arthur Johnson, who was not only my advisor, but a 
wonderful professor and mentor.  I would like to thank him for his guidance and more 
importantly for his patience through my marriage and pregnancy.  I am also grateful to 
my committee members, Dr. Wheaton and Dr. Felton for their guidance and support.  
Furthermore, I would like to acknowledge and thank Dr. Richard McCuen for helping me 
with the modeling and statistics portion of this thesis, as well as his assistance in editing.  
I would like to thank my subjects, and most importantly, Will Scott for helping during 
testing, guiding me during the entire process, and helping me to work out the kinks of this 
project.  I would also like to thank everyone involved in the human performance lab, 
including all graduate assistants and interns. Special thanks to Gary Seibel for making 
sure the environmental chambers were always up and running for me.   
 I would like to thank my husband for supporting me both emotionally and 
financially during this time, and most importantly, I?d like to thank my new daughter 
Grace for giving me a reason to finish my degree.   
iv
Table of Contents 
 
Dedication..........................................................................................................................iii 
Acknowledgements............................................................................................................iii 
List of Symbols................................................................................................................. vii
List of Tables ...................................................................................................................viii 
List of Figures..................................................................................................................... x 
Chapter 1: Justification ....................................................................................................... 1 
Chapter 2: Literature Review.............................................................................................. 3 
2.1.  Physiological Response to Temperature and Humidity.......................................... 3 
2.1.1. Evaporation....................................................................................................... 3 
2.1.2.  Humidity Effects on Heat Loss........................................................................ 4 
2.2.  Heat Exchange in Respirators................................................................................. 4 
2.3.  Effect of Clothing and Respiratory Protection  on Heat Exchange ........................ 5 
2.4.  Temperature and Humidity as Determinants of User Acceptability....................... 7 
2.5.  Sensitivity of the face in response to temperature and humidity.......................... 11 
2.6.  Breathing effects due to environmental conditions .............................................. 12 
2.7.  Heat Stress Indices................................................................................................ 14 
2.8.  Research Surveys.................................................................................................. 16 
Chapter 3: Research Goal and Objectives ........................................................................ 21 
Chapter 4: Research Methods ........................................................................................... 22 
4.1.  Procedures............................................................................................................. 22 
4.1.1.  Equipment/Apparatus .................................................................................... 22 
4.1.2.  Environmental Testing Conditions ................................................................ 25 
4.1.3.  Orientation and Consent ................................................................................ 26 
4.1.4.  VO
2
max Graded Exercise Test ...................................................................... 26 
4.1.5.  65-70% VO
2
max Subject Testing.................................................................. 27 
4.2.  Method of Statistical Analysis.............................................................................. 30 
4.2.1.  Outlier Detection............................................................................................ 30 
4.2.2.  Model Structure ............................................................................................. 31 
4.2.3.  Calibration...................................................................................................... 32 
4.2.4.  Sensitivity Analysis ....................................................................................... 33 
4.2.5.  Assessing Model Prediction Accuracy .......................................................... 34 
4.2.6.  Sample Size Determination............................................................................ 35 
Chapter 5:  Research Results and Discussion................................................................... 37 
5.1.  Subject Demographics and Characteristics........................................................... 37 
5.2.  Sensitivity Analysis of Heat Index ....................................................................... 38 
5.3.  Performance Time ? Heat Index Model................................................................ 40 
5.4.  Sample Graph Explanation ................................................................................... 45 
5.4.1.  Rectal Temperature vs. Time......................................................................... 45 
5.4.2.  User Acceptability vs. Time .......................................................................... 46 
5.4.3.  User Acceptability vs. Mean Facial Skin Temperature ................................. 51 
5.4.4.  Mean Facial Skin Temperature versus Time ................................................. 55 
5.4.5.  Rectal Temperature vs. Mean Facial Skin Temperature................................ 56 
v
5.5.  Analysis of Factors Affecting C
1
.......................................................................... 58 
5.5.1.  C
1
versus Minute Volume .............................................................................. 58 
5.5.2.  C
1
versus SN .................................................................................................. 59 
5.5.3. C
1
vs. Respirator Familiarity........................................................................... 60 
5.6.  Collective PT-HI model........................................................................................ 61 
5.6.1.  Sensitivity Analysis of the PT-HI model (Equation 5-9)............................... 70 
Chapter 6:  Conclusions.................................................................................................... 72 
Chapter 7:  Recommendations for Further Research........................................................ 73 
7.1.  Equipment............................................................................................................. 73 
7.2.  Subject Effort........................................................................................................ 74 
7.3.  Research Methods................................................................................................. 76 
Appendix 1 (Subject Excel Data) ..................................................................................... 79 
Appendix 2 (Mean Facial Skin Temperature Data).......................................................... 81 
Appendix 3 (Acceptability Data) ...................................................................................... 87 
Appendix 4 (Rectal Temperature Data)............................................................................ 93 
Appendix 5 (Performance Time ? Heat Index Model) ..................................................... 99 
Appendix 6 (Survey Data) .............................................................................................. 100 
Appendix 7 (C
1
graphs) .................................................................................................. 102 
Appendix 8 (Research Surveys)...................................................................................... 107 
Appendix 9 (Acceptability Scales) ................................................................................. 114 
Appendix 10 (IRB 05-0245) ........................................................................................... 117 
Bibliography ................................................................................................................... 127 
vi
List of Symbols 
A - Acceptability 
BACS ? Breathing Apparatus Comfort Scale 
BTPS ? Body Temperature and Pressure Saturated 
C
1
? y-intercept 
C
2
? x-intercept 
C
3
? scale parameter 
C
4
? shape parameter 
CLQ ? Claustrophobia Questionnaire 
EI ? Extroversion/Introversion parameter 
ECG ? Electrocardiogram 
F ? calibration form parameter 
F
i
? factors that influence output variable O 
FT ? Facial thermal scale 
HI ? Heat Index 
n
H - normalized half width 
HRmax? maximal heart rate (beats/min) 
HSI
BH
 - Heat Stress Index of Belding and Hatch  
FT ? Facial Thermal Scale 
IRB ? Institutional Review Board 
JP ? Judging/Perceiving parameter 
MBTI- Myers-Briggs Type Inventory 
n ? sample size 
O
0
? value of O at F
i
OT ? Overall Thermal Scale 
PAPR ? Powered Air-purifying Respirator 
PERQ ? Perceived Effort and Reward Questionnaire 
PT ? Performance time 
R ? correlation coefficient 
RF- Respirator Familiarity 
RH ? Relative Humidity 
RI ? Reference Index  
RS ? Relative Strain Index 
RPE ? Rating of Perceived Exertion  
R
s
? relative sensitivity 
SCBA ? Self contained breathing apparatus 
S
e
? standard error of the estimate 
SN ? Sensing/Intuition parameter 
S
y
? standard deviation of the criterion variable 
TF ? Thinking/Feeling parameter 
2/

t - t statistic at designated alpha 
 - degrees of freedom  
V
e
? Minute ventilation (L/min) 
V
t
? Tidal volume (L/min) 
VO
2
? Oxygen consumption (ml/kg) 
vii
VO
2
max
 
- Maximal oxygen consumption (ml/kg) 
WBGT ? Wet Bulb Globe Temperature (?C) 
WGT ? Wet Globe Temperature (?C) 
i
Y -i
th
 predicted value of criterion variable 
i
Y - i
th
 measured value of Y 
YSI ? Yellow Springs International  
 
viii
List of Tables 
 
Table 4-1.  Critical Values R
c
for a sample size of 10??????????....?....................31 
 
Table 5-1.  Absolute and Relative Sensitivities for Temperature??????????....?...39 
 
Table 5-2.  Absolute and Relative Sensitivities for Relative Humidity????????....?..39 
 
Table 5-3.  Model results for the performance time-heat index (PT-HI) model where PT=C
1
e
-
C2*HI
?????????????????????????????..???????41 
 
Table 5-4.  Average of user acceptability (%) for each of the subjects at the four heat index 
conditions..?????????????????????????????????..47 
 
Table 5-5.  Acceptability (%) values computed from Equation 5-8 for 27?C, 28?C, 37?C, and 
55?C for a range of mean facial skin temperatures ??????????????????54 
 
Table 5-6.  Performance time model results for varying sample sizes??...????????62 
 
Table 5-7.  Performance time (min) predictions from Equation 5-9 for values of V
e
(L/min), SN 
(ratio), R (number of times respirator worn), and HI (?C).???????????????63 
 
Table 5-8.  Predicted performance time values (min) and observed average performance time 
values (min) for ten subjects..??????????????????????????.69 
 
Table 5-9.  Relative sensitivity analysis of the PT-HI model (Equation 5-9)..???????...71 
 
Table A1.1.  Subject Demographics???????????????????????...79 
 
Table A1.2.  VO
2
max data???????????????????????????..79 
 
Table A1.3.  Performance Time data???????????????????????..79 
 
Table A2.1.  Mean facial skin temperature in ?C (recorded every 30 seconds) for each subject at a 
heat index of 27?C??????????????????????????????..81 
 
Table A2.2.  Mean facial skin temperature in ?C (recorded every 30 seconds) for each subject at a 
heat index of 28?C??????????????????????????????..82 
 
Table A2.3.  Mean facial skin temperature in ?C (recorded every 30 seconds) for each subject at a 
heat index of 37?C??????????????????????????????..84 
 
Mean facial skin temperature in ?C (recorded every 30 seconds) for each subject at a heat index 
of 55?C???????????????????????????????????85 
 
Table A3.1.  Acceptability (%) data (recorded every 2 minutes) for each subject at a heat index of 
27?C????????????????????????????????????87 
 
ix
Table A3.2.  Acceptability (%) data (recorded every 2 minutes) for each subject at a heat index of 
28?C????????????????????????????????????87 
 
Table A3.3.  Acceptability (%) data (recorded every 2 minutes) for each subject at a heat index of 
37?C????????????????????????????????????88 
 
Table A3.4.  Acceptability (%) data (recorded every 2 minutes) for each subject at a heat index of 
55?C????????????????????????????????????88 
 
Table A3.5.  Acceptability (%)for each heat index (?C) condition over time (min) calculated from 
Equation 5-5?????????????????????????????????88 
 
Table A3.6.  Time (min) to reach acceptability levels (30%, 40%, 50%) at various heat indices 
calculated from Equation 5-6??????????????????????????..91 
 
Table A4.1.  Rectal temperature in ?C (recorded every 30 seconds) for each subject at a heat 
index of 27?C????????????????????????????????..93 
 
Table A4.2.  Rectal temperature in ?C (recorded every 30 seconds) for each subject at a heat 
index of 28?C????????????????????????????????..94 
 
Table A4.3.  Rectal temperature in ?C (recorded every 30 seconds) for each subject at a heat 
index of 37?C????????????????????????????????..96 
 
TableA4.4.  Rectal temperature in ?C (recorded every 30 seconds) for each subject at a heat index 
of 55?C???????????????????????????????????97 
 
Table A5.1.  Performance time (PT) values (min) for each subject at various heat indices (?C).100 
 
Table A6.1.  Perceived Effort and Reward Questionnaire Results???????????...100 
 
Table A6.2. Claustrophobia Questionnaire Results?????????????????..100 
 
Table A6.3.  Physical Activity Questionnaire Results????????????????..100 
 
Table A6.4.  Respirator User Questionnaire Results.  Frequency represents the number of times 
the respirator had been worn prior to taking part in the study?????????????..101 
 
Table A6.5.  MBTI Personality Type Questionnaire Results?????????????...101 
 
x
List of Figures 
 
Figure 2-1. Diagram of the climate box used by Nielson et al. (1987), and Gwosdow et al. 
(1989)????????????????????????????..?????.......8 
 
Figure 4-1.  Hose connected from one environmental chamber through the exercise chamber to 
the respirator mask of the subject ?...?????????????????...??....??22 
 
Figure 4-2.  Calibration curves of humidity sensor and commercial psychrometer for temperature 
and humidity????????????????????????????????...24 
 
Figure 5-1.  Functions (n=29) for the PT-HI model (Equation 5-3) for all ten 
subjects...??????????????????????????????????.42 
 
Figure 5-2.  Functions (n=29) for the PT-HI model (Equation 5-3) for subjects 001, 145, and 358 
including original data points??????????????????????????..43 
 
Figure 5-3.  Comparison of Subject 145 (n=29) and Subject 419 (n=29).  Model  functions and 
original data points are shown.  Subject 145 has a C
1
value of 140.7 and a C
2
value of 0.05516.  
Subject 419 has a C
1
value of 11.482 and a C
2
value of 0.002157?..??????????...44 
 
Figure 5-4.  Average rectal temperature vs. time for all four heat index values.  R
2
=0.9884 for 
HI=27?C, R
2
=0.9773 for HI=28?C, R
2
=0.9949 for HI=37?C, and R
2
=0.9948 for HI=55?C...?..46 
 
Figure 5-5.  Average user acceptability (%) vs. time (min) for each of the four heat index 
conditions.????????????????????????.?????????..48 
 
Figure 5-6.  Model functions (n=29) for predicting time to reach unacceptable levels of comfort 
of the respirator according to the heat index of the environmental condition for acceptabilities of 
30, 40, and 50%...?????????????????????.?????????..50 
 
Figure 5-7.  Average user acceptability (%) vs. average mean facial skin temperature (?C) for all 
four heat index values?..??????????????????.?????????..52 
 
Figure 5-8.  Plot of average acceptability vs. average mean facial skin temperature using data 
from Table 5-5 for the four heat index conditions (n=26)...?????.?????????..55 
 
Figure 5-9.  Average mean facial skin temperature (?C) vs. time (min) for all four heat 
index conditions?????????????????????.?????????..56 
 
Figure 5-10. Average rectal temperature vs. average mean facial skin temperature for all four heat 
index values as well as original data points?????????????????????57 
 
Figure 5-11.  C
1
(from Equation 5-3) vs. V
e
for female and male subjects (n=10)??????59 
Figure 5-12.  The relationship between the model coefficient C
1
(from Equation 5-3) and SN for 
all subjects (n=10)??????????????????????????????...60 
 
xi
Figure 5-13.  The relationship between the model coefficient C
1
(from Equation 5-3) and 
Respirator Familiarity for all subjects (n=10)????????????????????.61 
 
Figure 5-14.  Performance time (minutes) vs. V
e
(L/min) computed from Equation 5-9???...64 
 
Figure 5-15.  Performance time (minutes) vs. the sensing-intuition (SN) parameter computed 
from Equation 5-9.??????????????????????????????..65 
 
Figure 5-16.  Performance time (min) vs. the number of times of the respirator was worn during 
manual labor prior to the current study computed from Equation 5-9.?????????..?66 
 
Figure 5-17.  Performance time (minutes) as a function of the heat index (?C) of  the 
environmental condition computed from Equation 5-9.????????????????67 
 
Figure 5-18.  Predicted performance time (n=4) vs. observed performance time (n=4) at the four 
heat index conditions.  Each observed performance time is an average of the performance times 
for ten subjects????????????????????????????????68 
 
Figure A3.1.  Average acceptability (%) vs. time (min) for all heat index conditions.  Data taken 
from Table A3.5???????????????????????????????..91 
 
Figure A7.1.  C
1
vs. V
e
from Table A1.2.  for each subject (n=10)???????????..102 
 
Figure A7.2.  C
1
vs. Perceived Effort and Reward Questionnaire results from Table A6.1. for each 
subject (n=10)???????????????????????????????...102 
 
Figure A7.3.  C
1
vs. Claustrophobia Questionnaire results from Table A6.2. for each subject 
(n=10)???????????????????????????????????103 
 
Figure A7.4.  C
1
vs. level of physical activity results from Table A6.3. for each subject 
(n=10)???????????????????????????????????103 
 
Figure A7.5.  C
1
vs. Extroversion/Introversion personality characteristic from Table A6.5. for 
each subject (n=10)?????????????????????????????...104 
 
Figure A7.6.  C
1
vs. Sensing/Intuition personality characteristic from Table A6.5. for each subject 
(n=10)???????????????????????????????????104 
 
Figure A7.7.  C
1
vs. Thinking/Feeling personality characteristic from Table A6.5. for each subject 
(n=10)???????????????????????????????????105 
 
Figure A7.8.  C
1
vs. Judging/Perception personality characteristic from Table A6.5. for each 
subject (n=10)???????????????????????????????...105 
 
Figure A7.9.  C
1
vs. Respirator Familiarity from Table A6.4. for each subject (n=10)???...106 
 
1
Chapter 1: Justification 
 
Respirators are worn in the workplace in order to protect the individual from toxic 
airborne substances; however, respirators are worn in only about 20-30% (Harber et al., 
1991; Nielsen et al., 1987; Gwosdow et al., 1989; DuBois et al., 1990) of the appropriate 
circumstances due to several factors, such as ventilation, work limitation, subject 
discomfort, psychological effects, thermal loading, and cardiovascular changes.  The 
focus of this research study will be on two of these factors:  subject discomfort and 
thermal loading.   
 High facial skin temperature is a major source of discomfort while wearing a 
respirator and the most common way to alleviate this discomfort is to remove the mask, 
although most working conditions do not permit this to occur (Johnson et al., 1997).  
Firefighters, mine workers, toxic chemical disposal crews, and other persons who require 
the use of respirators at work are subjected to various ambient conditions that elicit a 
physiological response.  This physiological response may impact performance and, 
ultimately, the quality of work produced.   
 According to several studies, facial skin temperature increases about 1-2?C while 
wearing a respirator during resting conditions (Johnson et al., 1997; Fox and DuBois, 
1993; DuBois et al., 1990).  During the stressful situations that require the use of a 
respirator, these thermal effects are exaggerated, which makes it difficult for the wearer 
to determine their limitations; hence, performance is affected.  Previous studies have 
focused primarily on the ambient conditions as a predictor of user acceptability.  While 
2
this information is important, it is necessary to determine how this acceptability 
correlates to performance time.   
 This study focused on determining a correlation between combinations of 
humidity and temperature based on the National Weather Service Heat Index (HI) on 
performance time at a moderate work rate under various simulated inspiratory air 
temperatures and humidities.  Other physiological parameters such as rectal temperature, 
facial skin temperature, and user acceptability were evaluated under warm environmental 
conditions.  The temperature and humidity conditions necessary for thermal comfort and 
optimal performance while wearing a respirator were unknown; these relationships will 
help to facilitate improved respirator design.   
 Respirators have been studied extensively due to the technical features that impair 
work performance.  A model developed to determine the relationship between inspired 
air conditions and performance time will provide a useful tool for respirator users.  This 
information will enable an employer to determine where optimum performance is 
achieved under various warm ambient conditions.  Employers will know how long an 
employee can perform safely according to both physiological and psychological factors.  
Furthermore, manufacturers may consider upgrading respirator equipment by providing 
the respirator wearer with a supply of fresh air kept at the most favorable ambient 
conditions, or design the inner surfaces of the respirator such that they absorb heat from 
the skin.  Ultimately, respirator users will benefit due to increased comfort, easier 
breathing, and improved work efficiency. 
3
Chapter 2: Literature Review 
 
The following sections detail the known physiological responses to air 
temperature and humidity levels, as well as the effect of respirators on thermal discomfort 
and performance.  Studies conducted on the effects of ambient conditions on user 
acceptability are discussed.  Work rates discussed are represented as a percentage of 
VO
2
max, which is defined as an individual?s maximal aerobic capacity.  The method for 
determining VO
2
max is discussed in section 4.1.4.   
2.1 Physiological Response to Temperature and Humidity
2.1.1. Evaporation 
 
The primary means for heat dissipation during exercise is evaporation via sweat.  
This process accounts for about 80% of the total heat loss from a person?s body during 
exercise.  Under resting conditions, 20% of the body?s total heat loss is due to 
evaporation (Wilmore and Costill, 2004).  During exercise, the body?s metabolic activity 
increases, blood flow increases, and the heart rate increases.  Consequently, skin 
temperature increases, creating a positive temperature gradient with the surrounding air.  
When the surrounding air is unsaturated, evaporation occurs and heat is dissipated from 
the body.  When the surrounding air is warm and humid, problems arise that create a 
smaller gradient for evaporation to occur, which inhibits adequate heat loss from the 
body.   
 
4
2.1.2.  Humidity Effects on Heat Loss 
 
Humidity levels affect evaporation as well as the body?s perception of thermal 
stress.  When the air is humid, the air contains water molecules that decrease its capacity 
to accept more water due to the decrease in the concentration gradient.  Thus, when a 
person is participating in vigorous activity in humid conditions, evaporation is greatly 
reduced.  Low humidity results in a higher concentration gradient, and evaporation is 
maximized; however, if sweat is not produced in direct correlation with the evaporation 
of water from the skin, the skin may dry out. Thermal stress is often times associated with 
a person?s psychological state, and the discernment of thermal stress may change with a 
change in the environmental conditions (Wilmore and Costill, 2004).   
 
2.2  Heat Exchange in Respirators 
Heat exchange at the face occurs through radiation, conduction, convection, and 
evaporation.  While wearing a respirator, radiation, convection, and evaporation are 
inhibited.  High facial skin temperature is a source of discomfort while wearing 
respiratory protective devices, as reported in several studies.  The comfort levels at rest 
and during exercise are 34.5?C and 31?C, respectively (Fox and DuBois, 1993).  Above 
these facial skin temperatures, the respirator user reported increasing mask discomfort.  
 The rate of radiation is approximately proportional to the temperature difference 
between the skin and the inner wall of the respirator face piece, though radiation inside 
the face piece is negligible.  Evaporation through the wall of the respirator is minimal, at 
best.  As for conduction and convection, heat is transferred through hot air from the lungs 
and bronchi to the skin; however, much of this heat is captured inside the mask and is 
5
unable to move through the respirator to the environment.  Some heat is carried outside 
the mask with exhaled air (Fox and DuBois, 1993).   
 Although respirators have expiratory valves, expired air does not leave the mask 
without mixing in the mask and transferring heat to the mask wall under warm 
environmental conditions.  If air left with minimal resistance, little residual heat would be 
left inside the mask.  This is not the case, however, because the expired air does not form 
a narrow channel through the expiratory valve.  The pressure required to open the valve 
broadens the air stream, re-circulating some of the expired air throughout the mask (Fox 
and DuBois, 1993).  Consequently, heat builds up on the face and the person begins to 
sweat, which causes discomfort.  These effects are exaggerated with increased exercise 
and heat stress.   
2.3.  Effect of Clothing and Respiratory Protection  on Heat Exchange
In order to gain a greater perspective on thermal effects during exercise, several 
studies were consulted regarding the effects of heat stress while wearing protective 
clothing.  Clothing effectively inhibits performance in hot, humid conditions by limiting 
heat exchange from the body to the surrounding air. Impermeable clothing inhibits 
effective cooling from the body so the head becomes a critical site for heat loss (James et 
al., 1984);  however, wearing a respirator limits heat exchange from the face.  
Consequently, the effect of the clothing and the respirator limit total body heat loss, 
which increases discomfort, affecting an individual?s work performance. 
 Clothing adds work by increasing the metabolic cost of performance by adding 
weight and restricting movement.  In a study conducted by Nunneley (1989), results 
showed that clothing inhibited evaporation by creating a humid environment and 
6
inducing thermal strain.  In a study conducted by Payne et al. (1994), the relationship 
between heat production and heat dissipation when wearing protective clothing that 
inhibited evaporation contributed to a rise in mean skin temperature.  White et al. (1989) 
concluded that wearing protective clothing and respirators induced dangerous 
thermoregulatory stress to the subject at low intensities of exercise in a neutral 
environment.   
 In general, the combination of protective clothing and respiratory equipment 
results in a decrement in performance compared to the unencumbered individual.  In most 
industrial settings where respiratory protection must be worn, impermeable clothing may 
be required as well.  The combination of the two causes an increase in heart rate and core 
body temperature; these effects are often accompanied by a decrease in performance 
(White et al, 1989;White and Hodous, 1987; James et al., 1984).   
 In a study conducted by White et al. (1989), two ensembles were worn with a Self 
Contained Breathing Apparatus (SCBA).  The first consisted of light work clothing, and 
the second consisted of a two piece chemical protective suit.  Work was performed at low 
intensity (23% VO
2
max) under three warm environmental conditions (10.6?C, 22.6?C, 
and 34?C).  Overall, it was concluded that in regards to rectal temperature, differences in 
the clothing ensemble were not significant, but the effect of thermal environment was 
significant.  High rectal temperature (>39.0?C) was a parameter used to assess the 
subject?s ability to continue work.  Once an individual?s rectal temperature exceeded this 
value, work was terminated.  Furthermore, the thermal gradient of rectal temperature 
minus mean skin temperature was almost  reduced to zero while wearing the chemical 
protective suit under hot conditions.  This was determined as a critical factor in an 
7
individual?s ability to tolerate heat.  In a study conducted by White and Hodous (1987) 
the use of a SCBA and impermeable clothing resulted in a decrease of exercise tolerance 
time by as much as 95.6%.  Tolerance time is a parameter defined by White et al. (1989) 
as the time required to achieve one of the following criteria:  1) 90% of maximum heart 
rate, 2) rectal temperature of 39.0?C, 3) skin temperature equaling or exceeding rectal 
temperature, or 4) objective or subjective sign of severe discomfort or fatigue (dizziness, 
nausea, etc.).   
2.4  Temperature and Humidity as Determinants of User Acceptability 
The effects of temperature and humidity as predictors of performance time have 
not been extensively studied; however, user acceptability of respirators according to these 
variables has been the topic of several studies, some of which are discussed below.   
 Nielsen et al. (1987) studied six subjects, all dressed similarly, in various ambient 
air temperatures, mask air temperatures, and air humidities inside the mask to determine 
user acceptability during exercise.  The subjects exercised for 15 minutes on a cycle 
ergometer at different combinations of ambient air temperatures (7?C, 16?C, 25?C), mask 
air temperatures (22?C, 27?C, 33?C), and mask air humidities (61% and 86%).  Skin 
temperature, heart rate, and skin wettedness were monitored during testing.  
Thermocouples were placed on the skin of the upper lip and on the cheek, and 
electrocardiogram (ECG) electrodes were fixed to the chest to monitor heart rate.  User 
rating scales were used to assess acceptability of the mask and whole body conditions.   
 The results of the study found that a significant interaction existed between the 
combined lip temperature and mask acceptability.  A significant interaction also existed 
between mean facial skin temperature and mask acceptability.  A high facial skin 
8
temperature decreased the acceptability of the mask.  Furthermore, low and high (22?C 
and 33?C, respectively) mask air temperatures resulted in low acceptability, whereas 
moderate mask air temperatures (27?C and 30?C) were acceptable.  Mask acceptability 
was considerably lower when the mask air was warm and humid than when it was warm 
and dry.  As expected, a higher mask air temperature resulted in a higher lip temperature 
when the mask air was humid compared to being dry.   
 Gwosdow et al. (1989) reported the effects of thermal discomfort on acceptability 
of respirators at rest under  
different combinations of 
ambient air temperatures (25?C, 
30?C, 35?C), mask air 
temperatures (27?C, 30?C, 33?C), 
and mask air humidities (47% 
and 73%) via use of the climate 
box shown in Figure 2-1.  A half-
mask respirator was used that 
 covered the mouth, nose, and part of the cheeks.  The skin temperature was recorded 
with thermocouples on the forehead, upper lip, and cheek.  Ratings of discomfort, 
thermal sensation, acceptability, and difficulty breathing were recorded during testing.   
 The results of the study indicated that the face, in particular the cheeks and 
forehead, are the regions most sensitive to warm stimuli.  The relationships between 
acceptability, discomfort, and thermal sensation versus lip temperature were the basis for 
this study.   According to the results, respirator discomfort and thermal sensation of the 
Figure 2-1.  Diagram of the climate box used by 
Nielson et al. (1987), and Gwosdow et al. (1989). 
9
face increased with an increase in lip temperature.  Furthermore, acceptability decreased 
with an increase in lip temperature, especially above a lip skin temperature of 34.5?C.   
The acceptability of a respirator was dependent on mask air temperature and humidity.  
Air temperatures of 27?C and 30?C with either of the two humidity conditions were 
always 100% acceptable regardless of room conditions.  At room temperatures of 25?C 
and 30?C, the acceptability of the respirator decreased for a mask air temperature of 33?C 
or above.  Similarly, with 73% humidity, a mask air temperature of 33?C or above had a 
lower acceptability than with 47% humidity.  Subjective responses of respirator thermal 
sensation, discomfort, and acceptability correlated (p<0.05) to lip skin temperature.   
 DuBois et al. (1990) reported results similar to the previous studies mentioned.  
The study measured facial discomfort as a function of facial skin temperature while using 
respirators that utilized tidal breathing rather than continuous airflow.  In respirators that 
utilize tidal breathing, heat and water vapor delivered by expired air contribute to the 
sense of increased temperature and humidity of the face.  Skin temperature was recorded 
via a thermocouple attached to the left nasolabial fold inside the mask, and a rating scale 
was used to record subject discomfort levels.  In this study, the subjects remained at rest. 
 The results of the study indicated that skin temperature of the face increased 
within a few minutes of putting on the mask.  Furthermore, discomfort was correlated 
with an increase in skin temperature.  As the skin temperature increased by 2.0?C, 
discomfort increased 1 unit.   
 Similar to the previous study, Fox and DuBois (1993) studied the effect of 
evaporative cooling of respirators on skin temperature and overall thermal sensation and 
comfort.  Resting subjects with lip skin temperatures of 34?C or less reported that the 
10
mask conditions were comfortable and almost 100% acceptable.  The study by Fox and 
DuBois (1993) reported results indicating that a high facial skin temperature (above 
31?C) is a major source of discomfort.   
 Two masks were used by Fox and DuBois (1993):  an aluminum mask with 
inspiratory and expiratory valves and a modified Scott model 66 twin-cartridge respirator.  
The facial skin temperature was measured with a thermocouple attached to the nasolabial 
fold inside the mask.  The subjects were measured at rest and during submaximal 
exercise.  The major results from this study indicated that the threshold for the comfort 
zone of skin temperature inside the mask for resting subjects began at a lip temperature of 
34.5?C, and at about 31?C for exercising subjects.   
 Laird et al. (2002) considered the effects of wearing a respirator on heart rate, and 
facial skin temperature on user acceptability.  The study conducted was two-fold: a 
laboratory study and a workplace study.  In the laboratory study, a standard filter 
respirator was worn for the first 15 minutes of exercise in the first session, and in the 
second session, the respirator was worn for only the second half of the exercise.  The 
respirator covered the mouth and nose only, and the facial skin temperature was recorded 
via bead thermistors positioned on the cheek and upper lip.  In the workplace study, 
subjects were asked to simulate their work tasks without a respirator, and then asked to 
carry out the tasks while wearing the respirator.   
 In the laboratory study, the mean temperature of the lip decreased when the 
respirator was removed and increased when the respirator was put on.  Wearing the 
respirator did not have a significant effect on heart rate or on the temperature of the 
cheek.  
11
Selkirk et al. (2004) recently studied the limits of firefighters in warm and humid 
environments.  The relationship between tolerance time to the work while wearing a 
respirator and work rate at three different environmental temperatures (25?C, 30?C, 35?C 
at 50% humidity) while wearing protective clothing was the main focus of the study.  The 
results showed that tolerance time decreased in response to an increase in temperature 
regardless of the work rate.   
2.5.  Sensitivity of the face in response to temperature and humidity
In many industrial situations, workers are required to wear protective clothing, 
which effectively inhibits evaporation from the body, so the head becomes the primary 
outlet for heat escape.  In many situations where protective clothing is required, a 
respirator may also be required.  In this case, evaporation from the face is repressed, 
forcing heat to build up in the body.  White et al (1991) indicated that the most common 
cause of discomfort while wearing respiratory protection was excessive heat inside the 
devices.  According to Gwosdow et al. (1989), these thermal conditions may often lead  
to subjective fatigue and an increase in the number of mental errors.  Preliminary 
research in our lab indicates that out of 165 surveyed respirator users, more than 40% 
indicated the most common reason for removing the respirator or not wearing it at all is 
that the respirator caused them to become too hot and sweaty (unpublished results).   
The heat flux per unit area of the bare face, 104 W/m
2
,
is double that of the rest of 
the body, which is about 50 W/m
2
(DuBois et al., 1990).  While wearing a respirator, the 
heat flux diminishes, causing an increase in temperature of the skin under the mask 
(DuBois et al., 1990).  Whereas the mechanism for discomfort is unknown, it is 
acknowledged to be a function of facial skin temperature, and may be caused by thermal 
12
sensation, sweating and hydration, condensation of expired air, or cutaneous blood flow 
(Gwosdow et al., 1989; DuBois et al, 1990; Nielsen et al., 1987).   
 Under normal conditions, the whole body thermal sensation is proportional to the 
area-weighted mean skin temperature; the findings by Gwosdow et al (1989) indicated 
that this relationship changed when thermal conditions inside the respirator changed.  For 
example, at a room air temperature considered neutral (25?C), increasing the respirator 
thermal conditions (temperature and humidity) changed the whole body thermal sensation 
from neutral to warm.  Nielsen et al. (1987) reported similar findings.  According to 
Nielsen et al. (1987), the use of different air temperatures in the mask compared to the 
ambient air produces local thermal stimuli to the skin surface beneath the mask, changing 
both the heat exchange from the skin surface to the air inside the mask and the heat 
exchange from the skin area under the edges of the mask.   
 
2.6.  Breathing effects due to environmental conditions
In a study conducted by Louhevaara et al. (1984), the effects of three different 
respirators on pulmonary ventilation, oxygen consumption, and heart rate were examined 
during rest, submaximal work, and recovery in well-trained young healthy men.  The 
conclusions of the study indicated that the dead space within the respirator increased the 
concentration of carbon dioxide in the inspired air, stimulating an additional effort in 
breathing, which subsequently increased oxygen consumption and heart rate.  However, 
according to Johnson et al. (2003), air-purifying respirator use under a variety of work 
conditions did not stress the cardiovascular system.  Thus, this phenomenon can be 
readily disputed.   
13
In the studies discussed in the previous section conducted by Nielsen et al. (1987) 
and Gwosdow et al. (1989), the results indicated that local thermal strain interfered with 
respiratory heat exchange.  The perceived work of breathing became more difficult as the 
heat content of the inspired air increased.  This was the case for only the conditions in 
which a mask was worn, and breathing was most difficult in the condition where the 
mask air was humid.   
 Lekeux (1988) studied the effect of environmental conditions on the breathing 
pattern of ponies.  The ambient temperature and humidity varied only according to the 
daily environmental conditions. Temperature and humidity were recorded prior to 
exercise and combined together to form a unitless measurement of the environmental 
condition.  Measurements less than 85 indicated cold and dry conditions, while 
measurements greater than 85 indicated warm and humid conditions.  The results of the 
study indicated that the conditions did increase the frequency of breathing.  Turner et al. 
(1992) performed a similar experiment on human subjects.  The results of Lekeux (1988) 
indicated that the frequency of breathing decreased in the cool, dry conditions, and tidal 
volume increased in the warm, humid conditions.  The increase in tidal volume likely 
contributed to an increase in minute ventilation.   
 Johnson et al. (2005) performed a study to determine peak inhalation flow rates 
during strenuous exercise.  Flow rates were measured at 80-85% VO
2
max without a 
respirator, with a powered air-purifying respirator (PAPR) and at the conclusion of the 
VO
2
max test without a respirator.  Major conclusions of the study indicated peak flow 
rates of up to 359 L/min (BTPS) for both respirator conditions, with flow rates for the 
14
PAPR exceeding breathing flow rates.  Furthermore, peak flow rates of up to 579 L/min 
(BTPS) were observed at 100% VO
2
max.   
 
2.7. Heat Stress Indices
Several attempts have been made at trying to find heat stress indices that predict 
the effects of hot environments on human physiological conditions and performance.  
There are over a dozen heat stress indices available; however many of them utilize heat 
loss, sweat loss, rectal temperature, and other physiological parameters in their 
calculation.  For the purposes of this study, a heat stress index is required in order to 
relate temperature and humidity.  One of the most widely used is known as the wet bulb-
globe temperature (WBGT) index because it incorporates the effect of temperature, 
humidity, radiation, and air movement.  
 McCann and Adams (1997) correlated WBGT index with performance in 
competitive distance runners.  Their results indicated that optimal conditions, designated 
by the best performance times were difficult to predict with a linear model.  Instead, they 
found that a curvilinear relationship more accurately described the relationship between  
several combinations of hot and humid environments and physiological responses.  Their 
major results showed that, regardless of how high the WBGT index was, runners 
performed better at lower humidity levels than at higher humidity levels.  Thus, the same 
WBGT index may produce different results in performance if in one case, temperature is 
high and humidity is low, and in the other, if temperature is lower but humidity is high.  
 Klemm and Hall (1972) delved into the issue of the utility of heat stress indices 
on physiological strain.  In their study, 20 different combinations of dry bulb-wet bulb 
15
temperatures were selected.  The participants remained at rest as physiological 
measurements were recorded.  Their results indicated that indices with lower dry bulb 
temperatures and higher wet bulb temperatures induced lower strains than those with 
higher dry bulb temperatures and lower wet bulb temperatures.  The conclusion, then, 
was that temperature and humidity differently affected the physiological response of the 
individual.   
 Ramanathan and Belding (1973) performed a study to determine the utility of the 
WBGT.  The objective of their study was to evaluate the WBGT index under 
combinations of environmental conditions and work rate.  Their results proved that a 
given level of WBGT had meaning dependent on environmental conditions, but that 
higher levels on the WBGT scale do not necessarily signify greater strain than lower 
levels.   
 Pulket et al. (1980) performed a study to compare available heat stress indices in a 
hot-humid environment.  His results indicated that the wet globe temperature (WGT) 
index gave the best correlation with physiological strain related to heat loss through the 
skin.  The WGT index is a single reading of the Botsford wet globe thermometer that 
exchanges heat with the surroundings by convection, evaporation, and radiation in a 
manner similar to that of a sweating man.  However, the WGT may not be completely 
suitable as a field heat stress index, but it did give a higher correlation with physiological 
strain than did the WBGT index.  Several other indices were studied, including the 
Relative Strain (RS) index, Reference Index (RI), and the Heat Stress Index of Belding 
and Hatch (HSI
BH
).  These all correlated with composite physiological strain; however, 
16
these heat stress indices incorporated heat loss, clothing effects, and other non-
environmental factors.   
 In order to evaluate the physiological strain of an individual with the environment, 
it is necessary to determine a heat index that effectively combines the effects of 
temperature and humidity.  The National Weather Service (2005) has developed a heat 
index equation as a result of extensive biometeorological studies. The index is a measure 
of how hot it feels when relative humidity (RH) is added to the actual air temperature.  
However, the equation is only useful for temperatures of 27?C (80?F) or higher, and 
relative humidites of 40% or greater.   
 
2.8. Research Surveys
One of the objectives of this research was to model work performance time.  This 
includes fitting coefficients of the model and explaining the factors that influence the 
values of the coefficients.  Therefore, several surveys were administered to each of the 
subjects in order to explain why some of the subjects had a good fit with the performance 
time model while others did not exhibit a good fit.  Five surveys were administered to 
each individual at the completion of their four testing sessions:  Perceived Effort and 
Reward Questionnaire (PERQ) (Tremblay et al., 2002), Claustrophobia Questionnaire 
(CLQ) (Radomsky et al., 2001), Raffenbarger Physical Activity Questionnaire 
(Raffenbarger et al.,1978), Respirator User Questionnaire (unpublished research), and the 
Myers-Briggs Type Inventory (MBTI) (Culp et al., 2001).  All surveys are located in 
Appendix 8. 
17
The respirator could be perceived by some as a constraint, specifically, one that is 
claustrophobic.  The CLQ was selected to measure this aspect of the test.  The CLQ is a 
quantitative measure of a person?s tendency to feel claustrophobic in certain situations.  
Claustrophobia is the fear of enclosed spaces and can often be unpleasant for people who 
are unable to cope with the fear of what might happen in that enclosed space.  
Claustrophobia is a combination of two separate fears:  fear of suffocation and fear of 
restriction.  The CLQ consists of 14 questions that pertain to fear of suffocation and 12 
questions that pertain to fear of restriction.  The test demonstrates high test-retest 
reliability, consistency, good discriminant validity, and predictive validity (Radomsky et 
al., 2001).   Because of its strong predictive and discriminant validity, the CLQ may be 
used in a variety of research applications.   
 The coefficients of the model may depend on the physical conditioning of the 
subject.  Therefore, a test was selected to provide a measure of the subject?s usual level of 
physical activity.  The Raffenbarger Physical Activity Questionnaire was developed by 
Dr. Ralph Raffenbarger Jr. for disease epidemiology studies.  The questionnaire estimates 
the number of kilocalories people expend per week in both sports and leisure activities.  
The survey is easy to complete and consists of questions that pertain to the number of 
stairs climbed per day, number of city blocks walked per day, and any sports or activities 
performed, including frequency and duration.  The subject is asked to average their 
activity over the entire year, and the researcher performs simple calculations to convert 
the activity to total energy expenditure.   
 A person?s performance can be influenced by their familiarity with the equipment.  
Therefore, an index of mask familiarity was selected.  The Respirator User Questionnaire 
18
was developed by Dr. Arthur T. Johnson (unpublished research).  The questionnaire 
evaluates a person?s familiarity with the respirator, reasons for discomfort while wearing 
the respirator, their overall attitude toward respirator use, and their ability to perform 
work while wearing the respirator.   The questionnaire was shortened to contain only 
questions relevant to respirator use during the current study and was used to assess the 
individual?s familiarity with the respirator, as well as their overall attitude toward the 
respirator.   
 The MBTI has been used for over 60 years and over this time, it has become one 
of the most widely used psychological instruments.  According to Consulting 
Psychologist Press, Inc. (Mountain View, CA), the MBTI is completed by approximately 
2 million people per year.  Many schools and employers use the instrument as a means 
for profiling characteristics associated with different personality types.  These personality 
types have been characterized (Culp et al., 2001) by four scales with opposite poles: 
extraversion (E)-introversion (I), sensing (S)-intuition(N), thinking(T)-feeling(F), and 
judging(J)-perceiving(P) Based on these scales, 16 distinct personality types emerge.   
 The tendencies of each scale (SN,TF, JP, and EI) were represented by a 
fraction in order to reduce the number of predictor variables within the MBTI 
parameters.  The following equations represent each grouped variable (Koh, 2004): 

 EI = number of Extraversion positives ? Total number of questions for 
Extraversion-Introversion 

 SN = number of Sensing positives ? Total number of questions of Sensing-
Intuition. 
19

 TF = number of Thinking positives ? Total number of questions for Thinking-
Feeling. 

 JP = number of Judging positives ? Total number of questions for Judging-
Perceiving. 
 A person who is extraverted  (E) acts first and thinks later, feels deprived when 
cutoff from interaction with the outside world, is usually open to and motivated by 
people, and enjoys a wide variety of relationships.  A person who is introverted (I) thinks 
first then acts, requires private time to recharge, is motivated internally, and prefers one-
on-one communication. 
 A person who exhibits sensing (S) characteristics mentally lives in the present, 
uses common sense and creates practical solutions by instinct, has rich memory recall, 
and likes clear and concrete information.  An intuitive (N) person mentally lives in the 
future, uses their imagination and creates new possibilities by instinct, improvises best 
through theoretical understanding, and is comfortable with ambiguous, fuzzy data. 
 A person who exhibits thinking (T) characteristics instinctively searches for facts 
and logic in a decision situation, naturally notices work that must be accomplished, is 
able to provide an objective analysis easily, and accepts conflict as a natural part of 
relationships.  A person who exhibits feeling (F) characteristics instinctively employs 
personal feelings in decision situations, is naturally sensitive to people?s needs, naturally 
seeks popular opinions, and is unsettled by conflict. 
 A person who exhibits judging (J) characteristics plans in advance before taking 
action, is focused on task-related action, works best when keeping ahead of deadlines, 
and naturally uses targets and standard routines to manage life.  A person who exhibits 
20
perceiving  (P) characteristics is comfortable moving into action without a plan, likes to 
multitask, is naturally tolerant of time pressure and works best close to deadlines, and 
instinctively avoids commitments that interfere with flexibility and freedom.   
 Some subjects were able to tolerate the hot, humid conditions better than others.  
The physiological and psychological factors inferred from the surveys were used to 
explain variations in performance time with the environmental conditions.
21
Chapter 3: Research Goal and Objectives 
 
The goal of the proposed research project is to develop a mathematical model that 
correlates performance time with the National Weather Service Heat Index (HI) to 
provide manufacturers with a useful tool for designing masks with the most favorable 
characteristics that are appropriate for optimal performance at the environmental 
conditions (i.e. temperature and humidity) prevalent at the location of use.  
To meet this goal, this research will address the following three objectives: 
 Model the relationship between inspired air conditions (humidity and 
temperature) and performance time while wearing a respirator;   
 Evaluate the user acceptability parameter (Breathing Apparatus Comfort Scale, 
or BACS) and assess its relationship with facial skin temperature;   
 Model the relationship between performance time with several parameters that 
may serve as predictors of performance time at the four environmental 
conditions 
- Correlate performance time with rectal temperature, facial skin 
temperature, and user acceptability   
- Examine factors that may pre-determine an individual?s 
performance time for screening purposes 
 
22
Chapter 4: Research Methods 
 
This research consisted of three stages:  obtain subject consent and orientation, 
perform the VO
2
max graded exercise test, and complete 65-70% VO
2
max testing under 
three different conditions.  A total of 10 subjects were recruited for the study. 
4.1  Procedures
4.1.1.  Equipment/Apparatus 
 
4.1.1.1. Environmental Chambers 
In order to supply warm humid air to the respirator mask, two Kolpak 
environmental growth chambers (Integrated Development & Manufacturing Company 
Chagrin Falls, OH) were used.  One environmental chamber was maintained at 30?C and 
40% humidity.  The subject exercised in this chamber, while air from the other 
environmental chamber was drawn into the exercise chamber via a long hose that was 
connected to the respirator mask (Figure 4-1).  The temperature and humidity in this 
chamber changed for each testing session, whereas the temperature and humidity in the 
chamber used for exercise remained constant. 
Figure 4-1.  Hose connected from one environmental chamber through the 
 exercise chamber to the respirator mask of the subject. 
23
The walk-in chambers are modular and consist of several aluminum tongue and 
groove sectional panels that contain four inches of foamed in-place polyurethane for 
insulation.  Each chamber consists of temperature control, a humidifying system, and a 
dehumidifying system.  A digital display on the exterior of each chamber allows for 
programmed control of these systems.  The temperature control ranges from 1.7?C to 
46.1?C (35?F to 115?F) and the humidity control ranges from 0% to 99%.  
 
4.1.1.2. Temperature/Humidity Sensor 
The Taylor Series 1452/1455 temperature and humidity sensor (Speranza?s 
Weather House; Hendersonville, NC) used for the research consists of an LCD display 
and external sensor probe with a 10 foot fixed cable.  The sensor measures indoor 
temperatures of -5?C to 50?C (23?F to 122?F) and outdoor temperatures of -50?C to 70?C 
(-58?F to 158?F), and measures indoor humidity from 20% to 99%.  The resolution is 0.1 
degrees F/C and 1% for humidity.  Although the resolution is good and the calibration 
with a sling psychrometer (Belfort Instrument Company Baltimore, MD) is sufficient, the 
reaction time for displaying a change in humidity is relatively poor.   
The indoor-outdoor temperature and humidity sensor was calibrated using a sling 
psychrometer (Belfort Instrument Company Baltimore, MD) at temperatures ranging 
from 62?F to 93?F and humidity ranging from 44% RH to 78% RH.  The sensitivity of 
the apparatus was relatively low, requiring an average time of 2 minutes to accurately 
display the relative humidity.  An analysis of variance performed indicated no significant 
difference (p < 0.05) between the actual temperature and humidity recorded by the 
psychrometer and the temperature and humidity displayed on the sensor.  The R
2
24
Figure 4-2.  Calibration curves of humidity sensor and sling psychrometer for 
temperature and humidity (n=7). 
 
value for the temperature calibration was 0.99 and the R
2
value for the humidity 
calibration was 0.97 as seen in Figures 4-2 above.   
4.1.1.3. Powered Air Purifying Respirator (PAPR) with blower 
The respirator mask chosen for this research was a Breathe Easy Turbo tight-
fitting powered air purifying respirator (3M St. Paul, MN) with a belt-mounted blower 
and filtration unit; however the filtration was excluded from the study.  The complete 
system used in the research included a blower unit, respirator headpiece, breathing tube, 
and a 4.5V external power source.   
Johnson et al. (2005) published voltages as well as the corresponding flow rates 
produced, and percentages of the maximum flow rate while wearing the tight fitting 
powered air purifying respirator (PAPR).  Based on this research, at 4.5V supplied to the 
motorized blower, the average flow rate produced is 103.46 L/min, which is 94% of the 
maximum flow rate able to be sustained by the blower unit.   
 
Temperature Calibration 
y = 0.9346x + 4.5352
R
2
= 0.9966
0
10
20
30
40
50
60
70
80
90
100
50 55 60 65 70 75 80 85 90 95 100
Sensor (?F)
P
s
y
c
h
r
o
m
e
t
e
r
(
?
F
)
Humidity Calibration
y = 1.0555x - 3.5747
R
2
= 0.9756
0
10
20
30
40
50
60
70
80
90
30 40 50 60 70 80 90
Sensor (% RH)
P
s
y
c
h
r
o
m
e
t
e
r
(
%
R
H
)
25
4.1.2.  Environmental Testing Conditions 
 
During each testing session, the subject exercised in an environmental chamber 
that remained at 30?C and 40% humidity.  The respirator was supplied with the following 
four combinations of temperature and humidity: 
 27?C, 50% humidity 
 27?C, 70% humidity 
 32?C, 60% humidity  
 37?C, 70% humidity  
In order to compare the conditions according to one variable, each state was assessed by 
the heat index developed by the National Weather Service.  Regardless of the magnitude 
of the individual environmental factors, the physiological responses will likely be 
different at different heat index values.  Choice of the conditions was based upon the heat 
index to see if there is a difference in performance time as the conditions feel warmer.  
 The following equation calculates HI based on two environmental variables and 
was developed for temperatures greater than 27?C (80?F) and relative humidity greater 
than 40% (National Weather Service, 2005):  
HI = -42.379 + 2.04901523T + 10.14333127RH ? 0.22475541TRH ? 
 6.83783x10
-3
T
2
? 5.481717x10
-2
RH
2
+ 1.22874x10
-3
T
2
RH +                                
 8.5282x10
-4
TRH
2
? 1.99x10
-6
T
2
RH
2
(4-1) 
where HI  = Heat Index (?F); T = ambient dry bulb temperature (?F); RH = relative 
humidity (integer percentage).  All heat index values were converted to ?C.  
 
26
According to the above equation, each of the testing conditions may be converted to 
following heat index: 
 At 27?C, 50% humidity, HI = 27?C 
 At 27?C, 70% humidity, HI = 28?C 
 At 32?C, 60% humidity, HI = 37?C 
 At 37?C, 70% humidity, HI = 55?C 
 
4.1.3. Orientation and Consent 
 
Institutional Review Board (IRB) approval on IRB application #05-0245 was 
received on May 18, 2005 (Appendix 10).  Ten young, healthy, normally conditioned 
subjects were recruited for the study.  Of these subjects, approximately half were not 
familiar with wearing the respirator; therefore, each subject was fully advised as to the 
requirements of their participation.  The subjects read and signed the informed consent 
document and medical history questionnaire.   The orientation session provided the 
subject with a detailed description of their rights, and it provided the investigators with 
information about the subjects? health and ability to partake in vigorous activity.  Any 
demographic or experimental data collected corresponded only to a subject number and 
may not be traced back to the individual.   
4.1.4.  VO
2
max Graded Exercise Test 
 
All prospective participants performed a maximal oxygen consumption test on a 
motorized treadmill (model 15.0Q, Image, Logan, UT) in order to determine each 
subject?s maximal aerobic capacity.  This test was used to develop each subject?s work 
rate during each of the testing sessions and each subject?s critical end point values.  
27
Participants warmed-up and stretched for approximately ten minutes prior to the start of 
the test.  The mask used during the test was a half-mask equipped with one-way 
inhalation and exhalation valves (Hans Rudolph Inc., Kansas City, MO).  This apparatus 
was interfaced with a standard Fleisch pneumotach (Phipps & Bird, Richmond, VA) and 
mass spectrometer (Perkin-Elmer, Pomona, CA) to monitor continuous expired airflow.  
Heart rate measurements were assessed using a Polar S810i wireless heart rate monitor 
(Polar Electro Inc. Lake Success, NY).  
 In order to determine VO
2
max, the subject participated in a graded exercise test to 
exhaustion.  The work rate was adjusted every three minutes until the participant became 
fatigued, failed to display a rise in oxygen consumption in accordance with the increase 
in work rate, or reached a maximal heart rate as determined from the following equation 
(Johnson, 2004): 
ageHR = 220
max
 (4-8) 
where HR
max
=maximum heart rate (beats/min).  All subjects? VO
2
max data may be found 
in Table A1.2. in Appendix 1.   
4.1.5.  65-70% VO
2
max Subject Testing 
 
Each session was conducted at 65-70% of the participant?s maximal oxygen 
consumption using the motorized treadmill. All sessions utilized a tight-fitting powered 
air purifying respirator (PAPR), which covered the entire face including the cheeks and 
forehead and supplied air to the respirator via a motorized blower.  The advantage to 
using the PAPR is to ensure a constant flow of the humid air to the user.  The subject 
exercised at a constant work rate throughout all four sessions.  Furthermore, each of the 
subjects dressed similarly, wore the same mask throughout all sessions, and exercised at 
28
the same temperature and humidity within an environmental chamber.  The only 
changing variables were temperature and humidity within the respirator. Testing 
conditions were randomized for each subject prior to the commencement of the study. 
Before exercise, the subject completed a five minute warm-up on the treadmill 
followed by five minutes of stretching.  The subject donned the heart rate monitor, a 
rectal probe (YSI 423, Yellow Springs International, Dayton, OH) was self-inserted 10 
cm inside the sphincter for measuring core body temperature, and three surface 
temperature sensors (YSI 4400, Yellow Springs International, Dayton, OH) on the face.  
The accuracy of the YSI 423 rectal probe was ?0.2?C from -1 to 60?C, ?0.1?C from 25 to 
45?C and the accuracy of the YSI 4400 temperature sensors was ?0.2?C for a range of  
-41? to +105?C. A Yellow Springs International (YSI) Precision 4400 Series 
telethermometer was connected to the sensors to record instantaneous skin temperature 
measurements (Yellow Springs, OH).  The sensors were placed on the mid-forehead, 
right cheekbone, and upper lip under the left nostril.  Mean skin temperature was 
calculated based on an equation proposed by Gwosdow et al (1989) for a weighted mean 
of ten local skin temperatures.  The weightings for the forehead, cheek, and upper lip 
were 0.046, 0.012, and 0.012 respectively.  Dividing each by the sum of the three yields 
the following equation for mean skin temperature: 
)(171.0657.0
lchfoface
TTTT ++ = (4-9) 
where T
face
 is the surface area weighted mean skin temperature for the face in ?C, T
fo
 is 
the forehead sensor temperature in ?C, and T
ch 
is the cheek sensor temperature in ?C, and 
T
l
is the upper lip sensor temperature in ?C. 
29
After the subject donned the heart rate monitor, skin temperature sensors, and 
inserted the rectal probe, the subject dressed in standard military fatigues and tennis 
shoes.  A full face piece respirator was fitted to the subject for a comfortable, but snug fit.  
The inlet of the respirator was connected via a breathing hose to the motorized blower 
that supplied air to the user from an environmental chamber.  At the outlet of the 
respirator, a combined temperature/humidity sensor probe monitored temperature and 
humidity of the expired air.   
The subject began to exercise at a treadmill speed and grade set at a work rate 
below 65-70% VO
2
max; the speed and grade were increased for approximately 90 
seconds before the final speed and grade corresponding to 65-70% VO
2
max was reached.  
At this work rate, the subject exercised until he or she reached exhaustion.  While the 
measure of exhaustion is purely subjective, the human monitor during subject testing 
observed whether substantial effort had been achieved based on maximum heart rate, 
Rating of Perceived Exertion (RPE),
 
or whether the subject expressed signs of severe 
discomfort.  The same human monitor was present throughout each of the testing 
sessions.  After each testing session, the subject was given a 5 minute cool-down then 
asked to provide reasons for termination. 
Four scales were used to determine how the subject felt every two minutes.  The 
Rating of Perceived Exertion (RPE) scale was used to determine how difficult the subject 
felt the work was.  A low score of 6 indicated that the work was very, very light, and a 
high score of 20 indicated that the work was very, very hard.  The Breathing Apparatus 
Comfort Scale (BACS) was used to determine how comfortable the subject felt the 
respirator was.  A low score of 0 indicated that the mask was very, very uncomfortable, 
30
and a high score of 10 indicated that the mask was very, very comfortable.  The Facial 
Thermal (FT) scale was used to determine how warm the subject?s face felt.  A low score 
of 1 indicated that the subject?s face felt very cold, and a high score of 7 indicated that 
the subject?s face felt very hot.  The Overall Thermal (OT) scale was used to determine 
how warm the subject?s overall body felt.  This scale is identical to the FT scale.  All 
scales are found in Appendix 9. 
Temperatures and humidities of both chambers were recorded prior to each 
testing session and following each testing session.  Heart rate was recorded every 5 
seconds and downloaded to a computer using the Polar Precision Performance software.  
Facial skin temperatures and core body temperatures were recorded every 30 seconds. 
Chamber temperature and humidity, the RPE, BACS, FT, and OT were recorded every 
two minutes.   
For the model, the dependent variable is performance time, indicated by the 
amount of time the subject was able to exercise until he or she reached exhaustion.  
Performance time was modeled as a function of the HI to determine the effect of the 
respirator.  Since performance time was an individual subject?s termination time, it was 
very subjective.  Therefore, several physiological and psychological factors may explain 
the variability of performance time.  
 
4.2 Method of Statistical Analysis 
4.2.1. Outlier Detection 
 
The Dixon-Thompson outlier test is used to statistically decide if an extreme 
event can be considered as an outlier (McCuen, 2003).  The test is applicable for sample 
31
sizes as small as three.   In order to conduct the test, the data are ranked from smallest 
(X
1
) to largest (X
n
), and the test statistic R is computed according to the sample size.  For 
a sample size of 8 to 10, the following equation for R is used: 
	
	









=

2
1
XX
XX
R
n
nn
(4-10) 
where X
n-1
 is the next to largest sample value, X
2
is the next to smallest value. The 
following are critical values (per % level of significance) for a sample size of 10: 
Table 4-1.  Critical Values R
c
for a sample size of 10 
 Critical Value 
5% 2.50% 1%
0.472 0.528 0.59
If the computed R is greater than R
c
, then the null hypothesis that the data point is not an 
outlier is rejected, and the data point (X
n
) is considered to be an outlier. 
4.2.2. Model Structure 
 
4.2.2.1.Graphical Analysis 
 
Graphical analyses are a useful first step in modeling to understand the structure 
of data.  Graphical analyses provide information on the effects of the independent 
variables on dependent variables.  For the proposed research, an association was expected 
between performance time and the HI.  Similarly, a correlation was expected between 
facial skin temperature and acceptability.  This part of the modeling effort provided a 
qualitative assessment of the degree to which one variable may be used to predict another 
variable (McCuen, 1985).  Graphical analyses are useful for identifying possible extreme 
events (outliers), the degree of association between variables, and the form of the 
32
relationships.  Misrepresenting the form of a relationship can reduce the accuracy of 
predictions made based on a model developed from the data.   
 Data may take several forms, each of which may be affected by the occurrence of 
extreme events, or outliers.  The first step is to identify the presence of outliers, then to 
identify the form and type of relationship between the variables (McCuen, 1985).  If 
extreme events occurred, these data points may affect the general trend of the data.   
 The data may take a linear or nonlinear form and have different degrees of 
correlation.  The data may be positively or negatively correlated, and the degree of 
correlation differs depending on the shape, trend, and slope of the data.  Graphs aid in the 
selection of the model structure that most accurately represents the data and the physical 
processes being modeled. 
4.2.3. Calibration 
 
Linear models do not always fit data to an acceptable degree; additionally, they 
may provide irrational predictions when the intercept is negative.  Thus, numerical 
optimization is a method of fitting nonlinear functions.  Calibration may occur via 
analytical, numerical, or subjective optimization.  Once the graphical analyses have been 
performed and a model structure identified, calibration is a process that determines the 
coefficients by minimizing the sum of the squares of the errors.  This equation takes the 
following form: 

=
=
n
i
i
i YYF
1
2
)(min  (4-11) 
where 
i
Y is the i
th
 predicted value of the criterion variable, 
i
Y is the i
th
 measured value of 
Y, and n is the number of observations on the criterion variable.  Analytical optimization 
33
utilizes differential equations to derive the unknown coefficients from objective 
functions.  For example, the derivative of Equation 4-11 with respect to the unknowns is 
set equal to zero, and the coefficients are determined by solving a set of simultaneous 
equations.   For numerical optimization, on the other hand, the coefficients are 
determined using an iterative, but systematic process.  The advantage to using the 
numerical process as opposed to the analytical process is that it can be used with more 
complex model structures and for non-differentiable model forms. 
 
4.2.4.  Sensitivity Analysis 
 
A model is intended to reflect the physical processes from which the data were 
measured.  A sensitivity analysis is a useful tool for assessing whether or not the model is 
a rational reflection of the processes.  In order for the model developed to be applicable 
to determining responses on a larger scale, the sensitivity of the model should be 
analyzed.  In other words, how sensitive the model is to errors, how important each of the 
variables and coefficients are in affecting the output, and how sensitive the goodness of 
fit is.  The general definition of sensitivity can be described as a function that reflects the 
effects of inputs on the output of the model.  Differentiating the output variable with 
respect to each factor and including system responses yields the general equation for the 
relative sensitivity of output variable (O) with respect to the factors (F) (McCuen, 2003): 
ii
s
FF
OO
R
/
/
0


= (4-12) 
where 
0
O is the value of O at some specified level of 
i
F ,
i
F are the factors that 
influence O , and 
s
R is the relative sensitivity.  Given the sensitivity of input F
i
with 
34
respect to a variation in the output O of the system response, this equation determines the 
relative importance of the factor.  The quantity 
s
R from Equation 4-12, is an indication 
of the relative importance of the factor F
i
on predictions of O
i
.
4.2.5. Assessing Model Prediction Accuracy 
 
Model accuracy may be determined by examining the bias of the model, the 
standard error of the estimate, and the correlation coefficient.  Bias is a statistical measure 
of the systematic variation of the errors of prediction.  Positive bias exists if the model 
consistently overestimates the value being measured, and negative bias exists if the 
model consistently underestimates the value being measured.  The general equation to 
calculate bias is: 

==
===
n
i
n
i
ii
i e
n
YY
n
ebias
11
1
)(
1
(4-13) 
A biased model exists if X is greater than zero and often indicates an incorrect model 
structure.    
 The standard error of the estimate (S
e
) is the square root of the sum of squares of 
the errors divided by the degrees of freedom.   It is a measure of the nonsystematic error 
variation of the data.  If the relationship between two variables is strong, then the 
standard error of the estimate will be smaller than the standard deviation of the criterion 
variable (S
y
).  The ratio of S
e
/S
y
is a measure of the relative improvement of the accuracy 
of predictions over predictions made with the mean of the variable.  A low S
e
/S
y
is more 
acceptable than a value closer to one.   
 The correlation coefficient (R) is a measure of the degree of linear relationship 
between two variables.  This index measures the goodness of fit of the model.  R
2
may 
35
also be used as a measure of the accuracy of predictions made by the model; however, the 
standard error of the estimate is a better measure because it is valid for nonlinear models 
as well as linear models (McCuen, 1993).  The following equation is used to estimate R: 
5.0
2
1
1








	
	







	








=
y
e
S
S
n
n
R

(4-14) 
where v is the degrees of freedom, and n is the sample size. 
 
4.2.6.  Sample Size Determination 
 
Three factors are used to determine sample size, two of which must be selected by 
the researcher:  the tolerable error and the level of confidence.  It is important to choose 
these values carefully because a large confidence interval will indicate an imprecise 
measure of the mean.  In most situations, the level of confidence is set between 90-95% 
and the tolerable error is chosen arbitrarily by the researcher.  The following equation 
was used to determine if the correct sample size had been chosen once the variance was 
known (Ott and Longnecker, 2001): 
2
22
2/
)(
E
z
n



= (4-15) 
where n = sample size, z
Y/2
 = z test statistic, Z
2
= population variance, and E = W/2 where 
W is the tolerable error. 
 For the current study, the standard deviation (Z) was computed for each testing 
condition and an average of the computed standard deviations was taken.  This value was 
5.98.  The tolerable error, which is representative of the width of the confidence interval 
of the mean, was arbitrarily chosen.  If the width is too narrow and confidence level too 
36
high, a large sample size will be necessary.  Therefore, W = 6, and E = 3.  A 95% 
confidence interval use used, which corresponds to a z value of 1.96.  Therefore, the 
sample size necessary to be 95% confident that the population mean is contained in the 
interval, is: 
n = [(1.96)
2
*(5.98)
2
]/3
2
= 15 
Although 22 subjects are necessary for a confidence interval of 95% with a width of 5, 
due to time constraints, 10 subjects were chosen for participation in the study.  Therefore, 
using a sample size of 10, the value of the z test statistic is: 
z
Y/2
 = (10
.5
*3
2
)/5.98
2
= 0.79 
This z value corresponds to a confidence level of 78%.  In order to improve this value to 
95%, more subjects must be tested or the width of the interval (W) must be expanded to 
7.5.     
37
Chapter 5:  Research Results and Discussion 
 
5.1. Subject Demographics and Characteristics
Ten healthy individuals participated in the research study: three female and seven 
male.  A retrospective analysis of the sample size indicated that a sample size of 15 
would have resulted in a 95% level of confidence.  A sample size of 10 resulted in an 
88% level of confidence.  Sample size determination is discussed in section 4.2.6.  The 
average age of the subjects was 26 + 7.3 years, the average height was 67.7 + 2.2 inches, 
and the average weight was 165.9 + 32 lbs (Table A1.1 in Appendix 1).  The average 
VO
2
max of the subjects was 29.16 + 5.87 ml/kg/min (Table A2.1. in Appendix 1).   
 All tests were randomized for each subject prior to commencement of the study.  
Three females and seven males participated in the study with ages ranging from 21 to 40.  
Of the ten subjects, four were familiar with respirator wear during manual labor.  Three 
of these four subjects (001, 145, and 358) had performance times that followed the 
expected trend for each of the conditions.  In other words, their shortest performance time 
occurred at the warmest condition, and the longest performance time occurred at the 
coolest condition.  The other seven subjects had performance times that did not vary 
greatly as a function of the heat index.  Although some of the subjects? performance 
times did not indicate that the warm and humid conditions affected them more than the 
more neutral conditions, all subjects were able to discriminate between the warmest 
(37?C, 70% RH) and the coolest (27?C, 50% RH) conditions.  All subjects indicated that 
the warmest condition made breathing very difficult from the beginning and some felt as 
though they were suffocating, leading to termination.  For the coolest condition, all 
38
subjects indicated that breathing became difficult toward the end of exercise, but 
termination occurred due to fatigue, boredom, or leg and muscle pain.  For the other two 
conditions (27?C, 70% RH and 32?C, 60% RH) subjects terminated for a variety of 
reasons including leg and muscle pain, fatigue, boredom, overall body discomfort due to 
heat, and difficulty breathing.   
 
5.2. Sensitivity Analysis of Heat Index
The heat index played a central part in this modeling effort.  Therefore, it was 
important to understand how it functioned as a variable.  For this reason, a sensitivity 
analysis of the heat index was undertaken prior to its use.  The sensitivity of a model is 
useful for assessing the relative importance of the predictor variables and making error 
analyses.  A sensitivity analysis was performed on the Heat Index equation developed by 
the National Weather Service (Equation 4-1) in order to determine how sensitive the 
equation was to changes in both temperature and humidity.  The analyses were performed 
on temperatures ranging from 25?C to 40?C and humidities ranging from 40% RH to 
80% RH.  These bounds were selected as representative of normal test conditions.  Both 
absolute and relative sensitivities were computed for temperature and humidity.  The 
absolute sensitivity is the first derivative of the dependent variable with respect to the 
independent variable and is useful in error analyses.  The relative sensitivity (Equation 4-
12) is the percentage change in the dependent variable for a 1% change in the 
independent variable; therefore, it is a useful indication of the relative importance of the 
independent variable. 
39
 As shown in Tables 5-1 and 5-2, a 1% change in the temperature at 40?C and 70% 
RH produced a 6.5% change in the actual heat index.   A 1% change in relative humidity 
at 40?C and 70% RH caused a 2.3% change in the actual heat index.  Therefore, at these 
values of T and RH, T was almost three times more influential than RH. 
 This indicated that the commonly used heat index equation may not be the best 
option for analyses of data collected during testing.  The coefficients of the heat index 
equation placed more emphasis on temperature than on the humidity.   
Table 5-1.  Absolute and Relative Sensitivities for Temperature 
Temperature (?C)  
RH (%) 25 30 35 40
50 0.5824 1.4787 2.3719 3.2651 absolute 
0.5735 1.6179 2.8668 4.3203 relative 
60 0.8311 1.9061 2.9811 4.0561 absolute 
0.8123 2.0808 3.5949 5.3546 relative 
70 1.1861 2.4357 3.6853 4.9349 absolute 
1.1594 2.6593 4.4446 6.5156 relative 
Table 5-2.  Absolute and Relative Sensitivities for Relative Humidity
Temperature (?C)  
RH (%) 25 30 35 40
50 0.0275 0.2829 0.705 1.294 absolute 
0.0269 0.3095 0.8522 1.7122 relative 
60 0.0085 0.359 0.8699 1.5411 absolute 
0.0083 0.3919 1.049 2.0344 relative 
70 -0.0105 0.4352 1.0347 1.7881 absolute 
-0.0102 0.4751 1.2479 2.3609 relative 
The absolute sensitivities in Tables 5-1 and 5-2 were also useful to show the 
effect of errors in measurements of T and RH.  Mathematically, an error in T of ]T or in 
RH of ]RH can be used to compute the error in the heat index ]HI by: 
 
]HI = ^HI/^T * ]T (5-1) 
 ]HI = ^HI/^RH * ]RH    (5-2) 
40
 
Errors in values of a computed heat index were more sensitive to errors in recorded 
temperature than to errors in recorded humidity.  For example, at 40?C and 70% RH, an 
error of 2?C led to a change of 9.8?C  in the heat index.  At 40?C and 70% RH, an error 
of 2% RH led to a 3.5?C change in the heat index.  The analysis indicated that the Heat 
Index Equation was almost three times more sensitive to errors in temperature than to 
errors in relative humidity. 
5.3.   Performance Time ? Heat Index Model
In order to predict performance time according to the heat index of environmental 
conditions, a model was fit to the data collected from each subject using the numerical 
least squares technique.  All performance time data are in Table A1.3. in Appendix 1.  
Graphing the data indicated that the structure of the model was nonlinear and that 
performance time was negatively correlated with the heat index of the environmental 
condition.  Two coefficients were used to define the model, which had the following 
form: 
HIC
eCPT
*
1
2

= (5-3) 
 
where PT=performance time (min), C
1
and C
2
are coefficients, and HI=heat index (?C).  
C
1
was the coefficient that represented the magnitude of the predicted values, while C
2
was the coefficient that described the rate of decline of the function as HI increased.  
Relatively high values of C
2
indicated a steep decline while relatively low values of C
2
indicated a gradual decline.  Generally, the two coefficients were correlated such that an 
increase in C
1
occurred simultaneously with a decrease in C
2
.
41
 Because the model was a nonlinear model, analytical methods for fitting C
1
and 
C
2
would not suffice; thus, numerical least squares was the appropriate technique.  For 
this technique, the coefficients for the model were determined using an iterative, but 
systematic process, which resulted in minimization of the sum of the squares of the 
errors.   The process involved numerically computing the derivatives of the objective 
function (Equation 5-3) and iteratively solving for the point where the derivatives were 
equal to zero (McCuen, 2003). Model results may be found in Table 5-3. 
Table 5-3.  Model results for the performance time-heat index (PT-HI) model where 
PT=C
1
e
-C2*HI
.
Subject S
e
S
e
/S
y
R
2
C
1
C
2
001 4.901 0.5634 0.7884 71.6 0.0358
145 0.306 0.0262 0.9995 140.7 0.05516
358 1.807 0.2627 0.954 58.95 0.03066
359 1.864 1.197 0.001 22.23 0.001219
379 1.001 1.193 0.001 13.75 0.001092
419 0.606 1.061 0.001 11.482 0.002157
420 1.213 1.174 0.001 14.442 0.00167
405 1.459 1.039 0.001 11.079 0.007195
401 0.89 0.57 0.7832 21.808 0.006405
343 0.555 0.15 0.9888 3.188 -0.04652
Model accuracy may be determined by examining the bias of the model, the 
standard error of the estimate, and the correlation coefficient.  For the PT-HI model, all 
models resulted in a bias very close to zero, suggesting that the correct model structure 
had been chosen. 
 The ratio of S
e
/S
y
is a measure of the improvement of the accuracy of predictions 
over predictions made with the mean of the variable.  A low Se/Sy is more acceptable 
than a value closer to one.  Subjects 001, 145, and 358 had low values of Se/Sy, 
indicating that the model accuracy for these three subjects was good.  Because of 
sampling variation inherent to the measured data, five of the models were inaccurate as 
42
evidenced from the high Se/Sy values and low R
2
values.  These subjects had 
performance times that did not vary greatly with a change in the heat index of the 
environmental conditions resulting in flat curves and an R
2
value close to zero as shown 
in Figure 5-1.  Calculations of performance time across a range of heat indices for each 
subject are in Appendix 5. 
0
5
10
15
20
25
30
35
40
45
27 32 37 42 47 52
Heat Index (C)
P
T
(
m
i
n
)
PT (001)
PT(145)
PT(358)
PT(359)
PT(379)
PT(419)
PT(420)
PT(405)
PT(401)
PT(343)
 
Figure 5-1.  Functions (n=29) for the PT-HI model (Equation 5-3) for all ten 
 subjects. 
 
The square of the correlation coefficient (R
2
) is a measure of the degree of 
relationship between performance time and the heat index.  This index measures the 
proportion of variation explained by the model.  For the subjects whose model accuracy 
was good, the goodness of fit of the model was very close to one.  Subject 001 had an R
2
of 0.7884, subject 145 had an R
2
of 0.9995, and subject 358 had an R
2
of 0.954.  Figure  
43
5-2 shows the functions for each of the subjects whose model accuracy and goodness of 
fit were adequate.   
0
5
10
15
20
25
30
35
40
45
50
27 32 37 42 47 52
Heat Index (C)
P
e
r
f
o
r
m
a
n
c
e
T
i
m
e
(
m
i
n
)
PT(001)
PT(145)
PT(358)
001
145
358
 
Figure 5-2.  Functions (n=29) for the PT-HI model (Equation 5-3) for subjects 
 001, 145, and 358 including original data points.   
 
It was evident from the PT vs. HI data that sampling variation can be very 
significant and adversely affect the modeling.  The sampling variation was evident from 
comparing the performance times measured at 27?C and 28?C in Table A1.3 in Appendix 
1.  While these should not vary by much relative to the variation of the measurements 
made at 37?C and 55?C, for many subjects the PT differences for a 1?C change in HI 
were large.  This would suggest that future studies might include replications to quantify 
the effects of sampling variation of each heat index. 
 
44
0
5
10
15
20
25
30
35
40
45
50
27 32 37 42 47 52
Heat Index (C)
P
e
r
f
o
r
m
a
n
c
e
T
i
m
e
(
m
i
n
)
PT(145)
PT(419)
145
419
 
Figure 5-3.  Comparison of Subject 145 (n=29) and Subject 419 (n=29).  Model 
 functions and original data points are shown.  Subject 145 has a C
1
value of 140.7 
 and a C
2
value of 0.05516.  Subject 419 has a C
1
value of 11.482 and a C
2
value 
 of 0.002157.   
 
Figure 5-3 shows the relationship between a subject who was sensitive to the 
change in the heat index (Subject 145) and a subject who was insensitive to the change in 
the heat index (Subject 419).  The function for subject 145 indicates a large difference in 
performance time for a heat index of 27?C versus a heat index of 55?C, whereas that for 
subject 419 shows almost no difference in performance times between these heat indices.  
Furthermore, the rate of decline for each function is indicative of the 25 fold difference in 
C
2
values. 
 The shape of the PT-HI relationship reflected a number of characteristics about 
the subject on which the relationship was based.  A steep decline was not necessarily 
indicative of good subject data, as the rate of decline can reflect the attitude and/or 
45
physical capability of the subject; however, a flat curve generally indicated data 
insensitive to differences in the environmental condition.  For example, subject 419 
terminated each testing session except that at 55?C due to sore legs, not because the 
condition became difficult.   
 Due to subject motivation, muscle soreness from other activities, and other 
external factors, only three subjects fit the model well.  In order to obtain a better 
understanding for why this was the case, explanation of differences in C
1
was attempted 
using several subjective and objective parameters.  Furthermore, once the differences had 
been accounted for, a new PT-HI model (Equation 5-9) was developed as a function of 
the new factors and combined all of the subjects.   
 
5.4.Sample Graph Explanation
5.4.1.  Rectal Temperature vs. Time 
 
Rectal temperature varied linearly with time and increased with a similar slope at 
each heat index value.  Figure 5-4 shows how rectal temperature varied over time for all 
ten subjects.  Each of the four heat index conditions showed little variation from one 
another, which indicated that rectal temperature increased independently of the 
environmental conditions, but nonetheless showed good correlation with time.  Rectal 
temperature data for all heat indices are located in Tables A4.1. to A4.4. in Appendix 4. 
46
R
2
= 0.9773
R
2
= 0.9948
R
2
= 0.9949
R
2
= 0.9884
37.2
37.3
37.4
37.5
37.6
37.7
37.8
37.9
38
0246810
Time (min)
R
e
c
t
a
l
T
e
m
p
(
C
)
27, n=21
28, n=16
37, n=14
55, n=13
 
Figure 5-4.  Average rectal temperature vs. time for all four heat index values.  
 R
2
=0.9884 for HI=27?C, R
2
=0.9773 for HI=28?C, R
2
=0.9949 for HI=37?C, and 
 R
2
=0.9948 for  HI=55?C 
 
5.4.2. User Acceptability vs. Time 
 
User acceptability of the respirator was measured using the BACS scale (0-10) 
with 0 indicating very, very uncomfortable conditions and 10 indicating very, very 
comfortable conditions.  All comfort scales are located in Appendix 9.  The scale was 
converted to a percentage to measure the acceptability of each of the heat index 
conditions using the following method: 
 
(BACS score/10)*100        (5-4) 
 During each testing session, each subject expressed his or her level of discomfort 
every 2 minutes using the BACS scale.  This data is located in Tables A3.1. to A3.4. in 
47
Appendix 3.  Averages were taken for each heat index across all ten subjects.  Table 5-4 
shows these values for each of the four heat indices.   
Table 5-4.  Average of user acceptability (%) for each of the subjects at the four heat 
index conditions. 
 HI=27 HI=28 HI=37 HI=55 
Time (min) 
Acceptability 
(%) 
Acceptability 
(%) 
Acceptability 
(%) 
Acceptability 
(%) 
0 72 75 75 66
2 55 61 52 43
4 46 48 41 29
6 41 39 36 26
8 32 32 24 25
10 24 29 21
12 19
The subject acceptability data of Table 5-4 were plotted in Figure 5-5 to show how 
acceptability declined over time.  However, the effect of the heat index was very 
minimal, since the trend at each heat index was nearly identical.  At any time, differences 
in acceptability were evident, but the trend was not consistent.  For example, at a time of 
0, the order was 28?C and 37?C (tie), then 27?C, and then 55?C.  At a time of 8 minutes, 
the order was 27?C and 28?C (tie), then 55?C, then 37?C. 
 
48
0
10
20
30
40
50
60
70
80
90
100
0 2 4 6 8 10 12 14
Time (min)
A
c
c
e
p
t
a
b
i
l
i
t
y
(
%
)
27, n=7
28, n=6
37, n=5
55, n=6
 
Figure 5-5.  Average user acceptability (%) vs. time (min) for each of the four 
 heat index conditions.   
 
A model was developed to represent the relationship between acceptability and 
time.  A comfort level of 40% corresponded to a BACS score of 4, ?fairly 
uncomfortable.?  Any scores below this level indicated that the subject felt that the 
respirator was uncomfortable.  An equation was derived to predict the amount of time in 
minutes it took for a person to deem the respirator unacceptable based on the heat index 
of the environmental condition.  Graphical analyses such as in Figure 5-5 indicated that 
the time (T, minutes) to reach a condition of unacceptability varied as an exponential 
decay function with the acceptability index (A,%), and that the rate coefficient of the 
exponential function varied approximately linearly with the heat index (HI).  These 
trends suggested the following model: 
49
))**((
321
AHI
eT
 ++
= (5-5) 
where T=time (minutes), _
1
, _
2
, and _
3
are model coefficients, HI=heat index (?C), and 
A=acceptability (%).  Raw data (Table A3.5.) and functions (Figure A3.1.) showing the 
relationship between acceptability level and time for the four heat index conditions are 
found in Appendix 3.   
 The empirical coefficients _
1
, _
2
, and _
3
were fitted using numerical least squares 
analyses of measured values of T, A, and HI.  The data of Table 5-4 were used to fit 
values of _
1
, _
2
, and _
3.
 An initial analysis showed that the data for a heat index of 27?C 
did not follow the trend expected; specifically, some of the acceptability values for 27?C 
are lower than the values for 28?C at the same time.  The model, including constants took 
the form: 
))**00056.002784.0(39.3( AH I
eT
+
= (5-6) 
where T is the predicted value of the time to unacceptability (min), and A is the level of 
acceptability (%).  Equation 5-6 over-predicted measured times by 0.13 minutes.  The 
standard error estimate of Equation 5-6 was 0.94 minutes (standard error ratio = 26%).  
This suggested that the model provided excellent accuracy.  The correlation coefficient 
was 0.97 or 94% explained variance.  These goodness of fit statistics indicated that the 
model provided accurate estimates of the time to unacceptability.  Figure 5-6 shows the 
relationship of Equation 5-6 between the time to reach an unacceptable level according to 
the heat index of the environmental condition for three levels of percent acceptability. 
50
0
1
2
3
4
5
6
7
8
9
10
27 32 37 42 47 52
Heat Index (C)
T
i
m
e
(
m
i
n
)
30%
40%
50%
 
Figure 5-6.  Model functions (n=29) for predicting time to reach unacceptable 
 levels of comfort of the respirator according to the heat index of the 
 environmental condition for acceptabilities of 30, 40, and 50%. 
 
The model equations indicated that the respirator became uncomfortable at a level 
of 40% at 2.85 minutes for the heat index of 55?C, at 4.26 minutes for the heat index of 
37?C, at 5.22 minutes for the heat index of 28?C, and at 5.34 minutes for the heat index 
of 27?C.  Overall, the time to reach an unacceptable level of respirator comfort decreased 
with an increase in the heat index of the environmental conditions, indicating that there 
was an effect of the condition on respirator comfort. The time at which the respirator 
becomes uncomfortable will provide useful information for those who set policies on the 
length of time at which the performance of respirator users might decline substantially. 
 The time to reach an unacceptable level of respirator comfort was not considered 
to be representative of termination time.  The model may serve as a useful tool for both 
respirator manufacturers and employers to determine how long it takes for the respirator 
51
user to become uncomfortable in the environmental conditions.  This time may be 
indicative of deteriorating physiological function at the worksite and decreased work 
efficiency.  While some individuals may push through the discomfort and continue to 
perform, others may terminate or remove the respirator at this time.   
5.4.3. User Acceptability vs. Mean Facial Skin Temperature 
 
User acceptability was measured in the manner described in section 5.4.2.  In 
order to compare the results obtained from the current study with those in the literature, 
user acceptability was graphed against mean facial skin temperature for all ten subjects at 
each of the four heat index values.  Figure 5-7 shows this relationship.  Data for mean 
facial skin temperature is located in Tables A2.1. to A2.4. in Appendix 2.   An overall 
relationship between user acceptability and mean facial skin temperature was evident.  It 
was expected that at the higher heat index values, acceptability would decrease more 
quickly as facial skin temperature increased, but this was not observed; however, as mean 
facial skin temperature increased for all four conditions, acceptability decreased.   
 
52
20
30
40
50
60
70
80
33.6 33.8 34 34.2 34.4 34.6 34.8 35 35.2
Mean Skin Temp (C)
A
c
c
e
p
t
a
b
i
l
i
t
y
(
%
)
27, n=6
28, n=5
37,n=5
55,n=4
 
Figure 5-7.  Average user acceptability (%) versus average mean facial skin 
 temperature (?C) for all four heat index values. 
 
In previous studies (Gwosdow et al., 1989; Dubois et al., 1990; Fox and Dubois, 
1993), exercising subjects with lip skin temperatures of 31?C or less indicated that the 
condition was almost 100% acceptable and any temperature over this threshold became 
increasingly unacceptable.  In the current study, a mean facial skin temperature (a 
weighted mean of lip, cheek, and forehead temperatures) was used instead of the lip 
temperature since Gwosdow et al. (1989) indicated that the cheeks and forehead are the 
regions most sensitive to warm stimuli.   
 Dubois et al. (1990) fit a linear regression through the data points representing the 
relationship between acceptability and mean facial skin temperature.  A similar approach 
was taken in order to compare results obtained from the current research study to those 
53
obtained by Dubois et al. (1990).  Dubois et al. (1990) obtained the following equation 
for unacceptability of mask conditions: 
A=-16.6347+0.527*F        (5-7) 
where A=unacceptability (%), and F=mean facial skin temperature (?C) (left nasolabial 
fold). 
Both nonlinear and linear regression equations were fit to the data collected in the current 
study; however, the linear model produced the best results.  The model took the following 
form to describe acceptability of mask conditions: 
A=1362.5-0.2*HI-38.40*F        (5-8) 
where A = acceptability (%), HI = heat index (?C), and F = mean facial skin temperature 
(?C).  The standard error estimate of Equation 5-8 was 8.74 (standard error ratio 52%).  
This suggested that the model provided good accuracy.  The correlation coefficient was 
0.87 or 74% explained variance.  These goodness of fit statistics indicated that the model 
provided accurate estimates of acceptability as a function of the heat index and mean 
facial skin temperature. 
 Dubois et al. (1990) concluded that acceptability began to decrease and the 
respirator mask became uncomfortable above mean facial skin temperatures of 34.5?C.  
In the current study, the threshold of acceptability was defined as a value of 40%, which 
corresponded to a BACS score of 4 (?fairly uncomfortable?).  Table 5-5 shows that for all 
heat index values, the respirator became uncomfortable above facial skin temperatures of 
34.5?C.  Figure 5-8 illustrates this result.  Beyond this temperature, acceptability fell 
below 40%.  This result confirmed those obtained by Dubois et al. (1990). 
 
54
 
Table 5-5.  Acceptability (%) values computed from Equation 5-8 for 27?C, 28?C, 37?C, 
and 55?C for a range of mean facial skin temperatures. 
Skin Temp (?C) HI=27?C HI=28?C HI=37?C HI=55?C
32.9 98.6 98.58 98.4 98.04
33 94.76 94.74 94.56 94.2
33.1 90.92 90.9 90.72 90.36
33.2 87.08 87.06 86.88 86.52
33.3 83.24 83.22 83.04 82.68
33.4 79.4 79.38 79.2 78.84
33.5 75.56 75.54 75.36 75
33.6 71.72 71.7 71.52 71.16
33.7 67.88 67.86 67.68 67.32
33.8 64.04 64.02 63.84 63.48
33.9 60.2 60.18 60 59.64
34 56.36 56.34 56.16 55.8
34.1 52.52 52.5 52.32 51.96
34.2 48.68 48.66 48.48 48.12
34.3 44.84 44.82 44.64 44.28
34.4 41 40.98 40.8 40.44
34.5 37.16 37.14 36.96 36.6
34.6 33.32 33.3 33.12 32.76
34.7 29.48 29.46 29.28 28.92
34.8 25.64 25.62 25.44 25.08
34.9 21.8 21.78 21.6 21.24
35 17.96 17.94 17.76 17.4
35.1 14.12 14.1 13.92 13.56
35.2 10.28 10.26 10.08 9.72
35.3 6.44 6.42 6.24 5.88
35.4 2.6 2.58 2.4 2.04
55
Figure 5-8.  Plot of average acceptability vs. average mean facial skin temperature 
 using data from Table 5-5 for the four heat index conditions (n=26).   
 
The results indicated that respirator comfort decreased above a mean facial skin 
temperature of 34.5?C during exercise conditions, which confirmed the results of Dubois 
et al. (1990).  This information will be useful for employers who may choose to monitor 
facial skin temperature as an indication of respirator comfort and ultimately, work 
efficiency.   
5.4.4. Mean Facial Skin Temperature versus Time 
 
Figure 5-9 shows the measured data of the mean facial skin temperature as a 
function of time.  The data are found in Tables A2.1. to A2.4. in Appendix 2.  The curves 
show an initial plateau or decline in facial skin temperature followed by a linearly 
increasing trend.  The plateau or decline did not represent significant variation, but the 
increasing trend was meaningful.  A substantial difference between each of the four 
56
conditions was not evident; however, the general relationship indicated an increase in 
facial skin temperature over time.  At 37?C, the slope in the linear portion of the data was 
much higher than that of 55?C.  This could be a result of lower performance times during 
testing at 55?C, during which the mean facial skin temperature never reached a peak. 
 
33.50
33.70
33.90
34.10
34.30
34.50
34.70
34.90
35.10
024681012
Time (min)
M
e
a
n
F
a
c
i
a
l
S
k
i
n
T
e
m
p
(
C
)
27, n=22
28, n=16
37, n=17
55, n=13
 
Figure 5-9. Average mean facial skin temperature (?C) vs. time (min) for all four 
 heat index conditions. 
 
5.4.5. Rectal Temperature vs. Mean Facial Skin Temperature  
 
Figure 5-10 shows the fitted relationships between the mean facial skin 
temperature and rectal temperature.  The graph was of value in characterizing the rate of 
increase of rectal temperature in response to the environmental stimulus.  At 27?C, the 
rectal temperature correlated to mean facial skin temperature (p<0.05) with an R
2
value 
57
of 0.9775.  At 28?C, rectal temperature correlated to mean facial skin temperature 
(p<0.05) with an R
2
value of 0.9139.  At 37?C, rectal temperature correlated to mean 
facial skin temperature (p<0.05) with an R
2
value of 0.9472.  At 55?C, rectal temperature 
correlated to mean facial skin temperature (p<0.05) with an R
2
value of 0.9395.   
As the heat index increased, the temperature range decreased.  This was mainly 
due to the fact that at the higher heat index values where performance time was shorter, 
less data were collected.  At a heat index of 55?C, the slope of the trend was .67.  At heat 
index values of 37?C and 28?C, the slope was .31.  At a heat index of 27?C, the slope was 
.47.  This indicated that mean facial skin temperature and rectal temperature increased 
independently of the environmental condition. 
R
2
= 0.9395
R
2
= 0.9472
R
2
= 0.9775
R
2
= 0.9139
37.2
37.3
37.4
37.5
37.6
37.7
37.8
37.9
38
33.6 33.8 34 34.2 34.4 34.6 34.8 35
Mean facial skin temperature (C)
R
e
c
t
a
l
t
e
m
p
e
r
a
t
u
r
e
(
C
)
27, n=24
28, n=18
37, n=14
55, n=13
 
Figure 5-10. Average rectal temperature vs. average mean facial skin temperature 
 for all four heat index values as well original data points.   
 
58
5.5. Analysis of Factors Affecting C
1
Parameter C
1
from Equation 5-3 was found to vary among data representing 
different subjects and test conditions.  Several surveys (section 2.8) were administered to 
the subjects and analysis of these surveys resulted in numerical values that varied 
according to the responses of the subject.  Responses to the surveys are in Tables A6.1. to 
A6.5 in Appendix 6.  In order to determine variants on which C
1
depended, a series of 
graphical procedures was undertaken.  These plots (Figures A7.1. to A7.9) are in 
Appendix 7.  A relationship was not evident between C
1
and all survey parameters; C
1
did have a relationship with minute volume, the sensing-intuition (SN) personality 
characteristic, and respirator familiarity.  The relationship between C
1
and minute volume 
was the first of these.   
5.5.1. C
1
versus Minute Volume 
 
Minute volume for each of the subjects is located in Table A1.2. in Appendix 1.  
This value represents the minute volume recorded for each subject at the work rate 
corresponding to 65-70% of their VO
2
max as determined during the VO
2
max test.   
 Figure 5-11 shows the relationship between the C
1
and minute volume (V
e
).  
Minute volume was essentially a measure of a person?s capacity to breathe air, and it was 
measured in liters of air inspired per minute.  Larger individuals and those individuals 
who were very physically active typically had higher minute volumes than those who 
were smaller and less physically active.  Figure 5-11 indicates that females typically had 
lower minute volumes than the men at their particular work rate as a percentage of 
59
VO
2
max during the study, and that a slight relationship existed between C
1
and V
e
for 
the men.  The data shown in Figure 5-11 suggested an increase in C
1
with increasing V
e.
 
0
20
40
60
80
100
120
140
160
0 1020304050607080
Ve (L/min)
C
1
Female
Male
 
Figure 5-11.  C
1
(from Equation 5-3) vs. V
e
for female and male subjects (n=10). 
5.5.2.  C
1
versus SN 
 
Figure 5-12 shows the relationship between C
1
and the Sensing-Intuition (SN) 
parameter from the MBTI assessment.  The results indicated that those with higher C
1
values also exhibited a lower tendency to have sensing characteristics, and instead had 
higher intuitive characteristics.  
 According to Figure 5-12, having more intuitive characteristics should increase 
the propensity of that individual to fit the initial PT-HI model (Equation 5-3); however, 
the opposite result was expected.  When modeled collectively (Equation 5-9), C
1
and SN 
60
showed a positive relationship instead of a negative one.  
 
0
20
40
60
80
100
120
140
160
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
SN
C
1
Figure 5-12.  The relationship between the model coefficient C
1
(from Equation 
 5-3) and SN for all subjects (n=10). 
 
The negative relationship of this bivariate plot was misleading because other more 
dominant variables also influenced the value of C
1
.
5.5.3. C
1
vs. Respirator Familiarity 
 
Respirator familiarity was quantified as the number of times an individual had 
worn the respirator while performing manual labor prior to the research study.  Figure 5-
13 shows the relationship between C
1
and respirator familiarity.  As an individual became 
more familiar with the respirator, they were also more inclined to have a better fit with 
61
the PT-HI model, which was indicated by a higher C
1
value.  
0
20
40
60
80
100
120
140
160
0 10 20 30 40 50 60 70 80 90
# of times Respirator worn
C
1
Figure 5-13.  The relationship between the model coefficient C
1
(from Equation 5-
 3) and Respirator Familiarity for all subjects (n=10). 
 
5.6. Collective PT-HI model
In reviewing the fitted coefficients of the original model (Equation 5-3) and 
knowing the subjects, it was evident that variations of the coefficients were related to 
characteristics of the subject.  In order to assess this variation, a number of parameters 
were analyzed, three of which showed strong correlations with the original model 
coefficients.  These parameters included psychological type, respirator familiarity, and 
minute volume.   
 The performance time-heat index model was modified to include the sensing-
intuition (SN) psychological parameter, the number of times the respirator was worn 
62
prior to the study (RF), and the minute volume (V
e
).  This collective model accounted for 
the variations in performance times among each of the subjects and provided accurate 
reproductions of the performance time. 
 In order to predict performance time according to the heat index of environmental 
conditions, an empirical model was fit to the data collected from each testing session 
using the numerical least squares technique.  Graphing the data indicated that the 
structure of the model was nonlinear and that performance time was negatively correlated 
with the heat index of the environmental condition.  A significant amount of variation 
existed between the subjects as evidenced from the differences between C
1
from Equation 
5-3.  Thus, in order to obtain an accurate model for performance time, several parameters 
were evaluated using surveys to determine how each affected the performance time.  The 
parameters chosen that affected C
1
were minute volume (V
e
), the Sensing-Intuition 
characteristic (SN), the number of times the respirator was worn (RF), and the heat index 
(HI).  The general form of the model is as follows, and model results are in Table 5-6: 
HIRFSN
e
eeeVPT
***
1
5432






= (5-9) 
where PT=performance time (min), Y
1
, Y
2
, Y
3
, Y
4
, and Y
5
are model coefficients, 
V
e
=minute volume (L/min), SN=sensing-intuition parameter, RF= number of times the 
respirator was worn, and HI=heat index (?C) 
Table 5-6.  Performance time model results from Equation 5-9.   
n e/y S
e
S
e
/S
y
R Y 1 Y 2 Y 3 Y 4 Y 5
40 -0.004 4.15 0.6141 0.8161 14.86 -0.1243 1.293 0.01624 -0.0198
(n= sample size, e/y = bias, S
e
= standard error of the estimate, S
e
/S
y
= prediction accuracy, R = correlation 
coefficient, Y
1
, Y
2
, Y
3
, Y
4
, and Y
5
are model coefficients)  
 In reviewing the residuals of the model, the prediction accuracy and goodness of 
fit statistics were improved over those shown in Table 5-3 from Equation 5-3.  While the 
63
goodness of fit statistics were not as good as they could be, they indicated that a 
collective model to include all subjects as well as other significant parameters was a good 
representation of the PT-HI process.   
 Some of the subjects had performance times that did not vary greatly with a 
change in the heat index of the environmental conditions.  Initially, the majority of 
subjects had poor goodness of fit statistics, with Se/Sy values above 1.0 and correlation 
values close to zero.  However, when the subjects were modeled collectively as a 
function of V
e
, SN, RF, and HI, the explained variation rose substantially.   
 The individual effects of each variable are shown in Table 5-7.  Performance time 
was calculated over a range of each variable while holding the other variables at mean 
values.   
Table 5-7.  Performance time (min) predictions from Equation 5-9 for values of V
e
(L/min), SN (ratio), RF (number of times respirator worn), and HI (?C) 
Ve PT (min)  SN PT (min)  RF PT (min)  HI PT (min) 
20 16.08 0.2 8.47 0 8.73 25 19.51
30 15.29 0.3 9.64 10 10.27 30 17.67
40 14.75 0.4 10.97 20 12.08 35 16.01
50 14.35 0.5 12.49 30 14.21 40 14.50
60 14.03 0.6 14.21 40 16.72 45 13.13
70 13.76 0.7 16.17 50 19.67 50 11.89
80 13.53 0.8 18.41 60 23.13 55 10.77
0.9 20.95 70 27.21
80 32.01
To show the effect of each variable, the model (Equation 5-9) was used to 
develop plots of each variable.  The values of the four variables were varied over the 
approximate range of the measured data.  The range of the computed values of 
performance time shown in the graphs suggested the importance of that variable.  Figures 
5-14 through 5-17 include estimations of performance time for each parameter listed in 
Table 5-7. 
64
13
14
14
15
15
16
16
17
20 30 40 50 60 70 80
Ve (L/min)
P
e
r
f
o
r
m
a
n
c
e
T
i
m
e
(
m
i
n
)
Figure 5-14.  Performance time (minutes) vs. V
e
(L/min) computed from Equation 
 5-9. 
 
Performance time decreased with an increase in an individual?s minute volume 
(V
e
)
.
This may be explained by differences in size, sex, and level of physical activity.  
Minute volume is the amount of air an individual inhales per minute.  Typically, a person 
who is of smaller stature has a lower minute volume since their lung capacity and need 
for air are smaller.  Furthermore, larger individuals may have different heat transfer 
characteristics from their bodies than someone who is of smaller stature.  Thus, someone 
with a smaller minute volume probably is not going to feel the effects of an extreme 
environmental condition as intensely as a person who inhales a great deal of air per 
breath.   
65
 
0
5
10
15
20
25
0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9
SN (ratio)
P
e
r
f
o
r
m
a
n
c
e
T
i
m
e
(
m
i
n
)
Figure 5-15.  Performance time (minutes) vs. the sensing-intuition (SN) 
 parameter computed from Equation 5-9. 
 
When modeled collectively in Equation 5-9, SN showed a positive relationship 
with performance time.  Performance time increased with an individual?s propensity to a 
higher SN ratio as determined from the MBTI assessment.  A person with a high ratio of 
SN possessed more sensing (S) characteristics as compared to intuition (N) 
characteristics.  This individual is likely to deal in facts and has an attitude that places 
emphasis on the present, rather than the future.  This individual tends to focus on 
activities at hand.  Because the N-type person is generally a daydreamer, they lack focus 
on activities in the present.  Therefore, performance time was expected to increase with 
an increase in values of SN, and the model coefficient related to SN showed the expected 
trend.  This relationship seems rational, since a person who thinks about current activities 
66
is more likely to stay focused on their performance and tends to have a longer 
performance time than a person who is already thinking about their next task. 
 
0
5
10
15
20
25
30
35
0 10 20 30 40 50 60 70 80
# of times respirator worn
P
e
r
f
o
r
m
a
n
c
e
T
i
m
e
(
m
i
n
)
Figure 5-16.  Performance time (min) vs. the number of times of the respirator was worn 
during manual labor prior to the current study computed from Equation 5-9. 
 
Each of the subjects was asked to disclose the number of times he or she had 
previously worn the respirator while performing manual labor prior to participation in the 
current study.  This number was used to quantify an individual?s level of familiarity with 
the respirator.  According to the results from Figure 5-16, an increase in respirator 
familiarity increased a person?s performance time at any given environmental condition.  
 An individual?s confidence in a test condition directly influences their 
performance.  An individual?s familiarity with the respirator mask imparts confidence in 
their ability to perform while wearing a mask.  Respirator masks are uncomfortable due 
67
to the design components necessary for adequate protection; a person who is familiar 
with this discomfort likely performs better than a person who is not familiar with the 
respirator.  The novice can be overwhelmed by the discomfort and has uncertain 
expectations.  This lack of confidence likely leads to a lower level of performance.  
Someone who has worn the respirator repeatedly while performing physical labor will 
likely push through the discomfort and only terminate because of the limitations the 
environmental condition placed upon them.   
 Performance time decreased with an increase in the heat index of an  
 
environmental condition.  The decay coefficient of the heat index component of the  
 
model exhibited the expected direction.  
0
5
10
15
20
25
27 32 37 42 47 52
Heat Index (C)
P
e
r
f
o
r
m
a
n
c
e
T
i
m
e
(
m
i
n
)
Figure 5-17.  Performance time (minutes) as a function of the heat index (?C) of 
 the environmental condition computed from Equation 5-9. 
 
Figure 5-17 shows the range of PT for a range of heat indices while holding the 
other variables at their means.  This result was expected since exercising in a very warm, 
68
humid environment compared to one that was cooler and less humid stressed the subject 
on a number of levels, forcing them to terminate sooner.  Inside the respirator mask, 
evaporation was inhibited, leading to a build-up of saturated air from the respiratory 
system.  This increased facial skin temperature, making the subject feel unbearably 
warm.  Furthermore, breathing humid air in comparison to air that was less humid put a 
strain on the respiratory system, perhaps forcing labored breathing and a feeling of 
suffocation.  All subjects felt as though they were hyperventilating and were not getting 
enough air during the warmest condition, citing difficult breathing as the reason for 
termination.     
Figure 5-18.  Predicted performance time (n=4) vs. observed performance time 
(n=4) at the four heat index conditions.  Each observed performance time is an average of 
the performance times for ten subjects. 
 
69
 Figure 5-18 shows the trend for predicted performance times from the model 
(Equation 5-9) versus average values of the observed performance times.  The R
2
value 
was 1.0 indicating a good linear relationship.  The slope was 0.99, indicating that the 
model predicts performance time well.  Table 5-8 displays the observed and predicted 
values for performance time. 
Table 5-8.  Predicted performance time values (min) and observed average performance 
time values (min) for ten subjects. 
HI (?C) Predicted PT (min) Observed PT (min) 
27 18.1 18.1
28 18.5 18.5
37 15.3 15.3
55 12.3 12.3
The model (Equation 5-9) will be useful for manufacturers because it can help 
them to design respirators appropriate for the locations and environments under which 
they will be used. This may include attaching motorized blowers with more efficient 
filters to lower the humidity inspired by the user or designing masks with enhanced 
evaporative cooling.   
 Such a model would be useful for employers to determine the factors that 
contribute to the detriment of performance time as the inspired air conditions become 
warmer and more humid.  Employers may use the model for pre-screening purposes and 
for determining how long they can expect their employees to perform efficient work.  
Knowledge of an individual?s V
e
, familiarity with the respirator, and/or SN-type 
personality, can allow employers to gauge the performance time of workers under a 
variety of environmental conditions.   
 The PT-HI model (Equation 5-9) predicts performance time across a range of heat 
indices; however, the performance time values are all less than 20 minutes.  While an 
70
individual may not stop working after this time, they may remove the respirator, exposing 
them to toxic substances.  Employers must be aware that this is occurring, and thus they 
may choose to allocate their human resources more effectively.   
 Furthermore, the model would be useful in respirator training programs to help 
illustrate expected performance.  Ultimately, the model can result in increased comfort of 
respirators, better work efficiency, and less injury and heat related illness of respirator 
users.   
 The PT-HI model (Equation 5-9) should be used with some caution.  The R value 
was 0.82, meaning that performance time can not always be accurately predicted from 
V
e
, RF, SN, or HI.   An R value closer to 1.0 would be indicative of a better model; 
including more parameters to better explain the variation among subjects would likely 
achieve this.  However, the collective model (Equation 5-9) had better prediction 
accuracy than the original model that included only the heat index as evidenced in Table 
5-3.  This is an indication that data relative to individual subject characteristics must be 
collected in order to explain the performance of individuals performing work in warm, 
humid environments.   
5.6.1.  Sensitivity Analysis of the PT-HI model (Equation 5-9) 
 
The performance time model of Equation 5-9 included four variables, each 
representing a different characteristic of the system.  The four predictor variables 
represented the environmental conditions (HI) of the test, the experience (RF) of the 
subject, the physiological stature (V
e
) of the subject, and the psychological nature (SN) of 
the subject.  To assess the importance of these four factors, a relative sensitivity analysis 
71
of the PT-HI model (Equation 5-9) was made.  The four relative sensitivity functions are 
as follows: 
 ^PT/^V
e
* V
e
/PT = Y
2
(5-10) 
 ^PT/^SN * SN/PT = Y
3
SN
 
(5-11) 
 ^PT/^R * RF/PT = Y
4
RF
 
(5-12) 
 ^PT/^HI * HI/PT = Y
5
HI
 
(5-13) 
where the coefficients Y
2
, Y
3
, Y
4
, and Y
5
are the values given in Equation 5-9. 
 While the relative sensitivity functions can be evaluated at any level of a variable, 
the mean values are most commonly used.  Based on the data for the ten subjects of this 
research, the results of a sensitivity analysis are summarized in Table 5-8.  The results 
showed that the environmental (HI) and psychological (SN) factors were the most 
important, with relative sensitivities of more than 80%.  Experience with the respirator 
(RF) was less important, but a 1% increase in respirator use will lead to about a 0.5% 
increase in PT.  The physiological factor (V
e
) was the least important, but it was still 
important, with a relative sensitivity of about 30%.  The relative sensitivities appeared to 
give rational indications of the relative importance of the four predictors and general 
intent of the effects.   
Table 5-9.  Relative sensitivity analysis of the PT-HI model (Equation 5-9). 
Variable Mean 
PT Model 
Coefficient Model form 
Relative 
Sensitivity 
Function 
Relative 
Sensitivity 
SN 0.57 1.561 exponential Y 3SN 0.89 
HI 36.75 -0.0231 exponential Y 5HI -0.849 
R 29.65 -0.0183 exponential Y 4RF 0.542 
V
e
54.08 -0.315 power Y 2 -0.3165 
72
Chapter 6:  Conclusions 
 Performance time correlated strongly with V
e
, SN, RF, and HI.  Physiological 
variables such as rectal temperature and facial skin temperature did not vary 
significantly with the change in the environmental conditions.   
 A sensitivity analysis of the PT-HI model indicated that the most important 
predictors of performance were the heat index and personality type, followed by 
respirator familiarity, then minute volume.   
 Individual subject characteristics were more important than physiological 
variables as predictors of performance times.     
 The respirator became uncomfortable above a mean facial skin temperature of 
34.5?C 
 Times to reach various levels of acceptability at the environmental conditions 
were determined 
 
73
 Chapter 7:  Recommendations for Further Research 
 
7.1  Equipment
During the course of the research several problems with the environmental 
chambers arose.  Because they were several years old, the compressors and chillers failed 
repeatedly and this forced testing to be delayed for several months.  A solution to this 
problem would be to design a suitable environmental chamber that could supply warm 
humid air to the respirator mask and obtain equipment that would perform properly.   
 The humidity chamber could be constructed similar to the device (Figure 2-1) 
built by Gwosdow et al. (1989), with some modifications made to enhance the ability of 
the air to become saturated more quickly.  Furthermore, the device should be capable of 
supplying the subject with the appropriate flow rates conducive to moderate intensity 
exercise.  According to Wilmore (2004), the maximum peak flow rate of large men is on 
the order of 100 L/min, although the average flow rate is around 24 L/min. 
 A Plexiglas chamber would be filled with the volume of water needed to supply 
the specified humidity and air temperature within the chamber.  A heating component 
capable of controlling the temperature of the system would consist of several coils.  
Insulated tubing would connect the outlet of the humidity chamber with the inlet of the 
respirator mask in order to minimize heat losses.  A fan system should be used to push air 
into the chamber, which would pass through a series of grids on the lid of the chamber to 
allow the air to pass through slowly, and thus, allow for maximum saturation to occur.   
74
 The outlet of the chamber should be connected directly to an insulated hose, 
which in turn, connects to the inspiratory valve of the respirator.  A suitable system 
designed to deliver warm humid air to the respirator mask would be another suitable 
method for subject testing considering the equipment problems experienced with the 
commercial environmental chambers used for this research. 
 The temperature and humidity sensors used to monitor expired air inside the 
respirator masks were not sufficiently quick to display instantaneously the temperature 
and humidity.  However, the sensor did reflect a relative change in temperature and 
humidity over the course of the testing session.  A sensor capable of displaying breath by 
breath changes in temperature and humidity would be preferred in order to separate and 
calculate heat losses from the face and respiratory system.  
7.2  Subject Effort
Performance time was directly related to subject effort, which may be a function 
of several variables.  Since performance time was the independent variable in the model, 
it was necessary to attempt to explain the variables that may influence subject effort.  The 
inability to control the subject effort can introduce significant errors into the data 
collected.  Subject effort variables include, but are not limited to, subject motivation, 
familiarity with the respirator, level of physical fitness, emotional state prior to testing, 
and personality type.  If each of these variables could be controlled, a more accurate 
model could be developed.   
 Subject motivation may arguably be the strongest predictor for how well a person 
is able to perform work; however, it is also the hardest to quantitatively measure.  
75
Existing indices are purely subjective and may be a function of the person?s mood, 
overall attitude, physical health, stress level, and diet.  A questionnaire could be 
administered to measure a person?s motivation based on the aforementioned factors, but 
it is likely to change on a daily basis and thus must be administered before each testing 
session.  A quantitative score should be given to each subject for each testing session 
completed in order to measure their personal motivation for performing the work.  It may 
also be necessary to administer a second test immediately after testing to determine if the 
subject?s motivation has waned, as it may for an intuitive (N) personality type.   
 Four out of the 10 subjects tested for this study were unfamiliar with wearing the 
respirator.  In other words, they had worn a respirator less than 10 times while 
performing physical labor.  Because a respirator is uncomfortable to begin with, a person 
who is not familiar with this feeling may be more inclined to terminate because of this 
discomfort before the effects of the environmental condition can be felt.  Furthermore, 
one of the subjects tested was very familiar with the respirator.  He was a member of the 
armed forces for 7 years, and frequently donned the respirator for several hours at a time 
in extreme environmental conditions.  During this time, regardless of the discomfort, he 
was unable to remove the respirator and was forced to endure the condition.  Therefore, 
during testing he pushed through each of the conditions with relative ease and did not cite 
the conditions as his reason for termination. 
 The remaining five subjects had worn the respirator several times while 
exercising in previous studies in the lab and were, therefore, moderately familiar with the 
discomfort of wearing a respirator.  Attempting to quantify the familiarity with wearing a 
76
respirator may help to explain the performance time of each individual.  Therefore, 
research is needed to more accurately quantify respirator mask familiarity as a predictor.   
 A third factor that may be related to subject effort is the level of physical fitness.  
The body temperature of a person who is out of shape may increase more rapidly.  The 
person may become dehydrated more quickly and may be inclined to terminate sooner 
than someone who is more physically fit.  A quantitative measure of a person?s physical 
fitness level would help to explain differences in performance time under each of the 
testing conditions.  This measure may include the duration and frequency of exercise and 
type of exercise performed.   
7.3.   Research Methods
It is imperative to obtain an accurate representation of the population when 
performing a research study.  The number of subjects affects the accuracy of conclusions 
made from the measured data, with accuracy increasing with the sample size.  However, 
it is also important to ensure that subjects show variability in the important factors.  Both 
the sample size and the variability inherent to the data collected determine knowledge 
gained from the data analysis.  While only ten subjects were tested in this research, the 
subjects varied in physical fitness, personality type, aerobic capacity, and mask 
familiarity.  This enabled the research to identify important relationships.  However, 
more knowledge about the population could be gained by testing more subjects.  These 
subjects should include equal numbers of men and women, they should represent all 
levels of physical fitness, and represent both subjects who are familiar with wearing a 
respirator and those who are unfamiliar with wearing a respirator.  By including subjects 
with diverse characteristics the analyses are more likely to detect the true effects of the 
77
inputs.  Furthermore, in order to validate the model, more tests on the ten subjects who 
participated in the study are needed.   
Since each subject has a different VO
2
max, V
e
, and tidal volume (V
t
), it is 
possible that the amount of oxygen utilized and amount of air intake per minute could 
help to explain why some subjects terminate sooner than others at each of the testing 
conditions.  For example, some of the larger subjects seemed to be affected by the 
conditions more so than some of the smaller subjects.  This could be a direct function of 
how much air is being utilized per minute.  The larger subjects were perhaps feeling the 
warm humid conditions sooner and more intensely than the smaller subjects because they 
were breathing in a larger volume of air per breath.  If continuously expired air data were 
monitored during each testing session, this could provide the researcher with more 
information regarding the relationship between such parameters as minute volume (V
e
)
and tidal volume (V
t
) to performance time under the environmental conditions.   
In order to gain a better understanding of each subject?s attitude prior to each 
testing session, a state anxiety test could be administered.  This test would give the 
researcher information on how the subject is feeling that day, and a quantitative score is 
reported.  A low score indicates low anxiety and a high score indicates high anxiety.  
Since a person feels differently day to day, this score could provide some insight on why 
a subject performs particularly well on one day, and poorly on the next day, especially 
when the difference is not related to the change in the environmental condition.  Such 
inconsistency in performance introduces error into the measured data.  Such error could 
mask effects or suggest significant effects that are not accurate. 
78
 The performance time ? heat index model was developed from exercise in 
neutral-warm conditions.  In order to develop a more accurate model capable of 
explaining how performance time could be predicted for exercise in cooler environments, 
three to four cool, dry conditions should be added to the protocol, and the analyses 
repeated.  The model developed indicates a negative correlation between performance 
time and the heat index of the condition; however, a more accurate indication of the 
underlying function could be obtained with a wider variation in the data.  A parabolic 
function may be observed if the cooler conditions were included in the protocol. 
79
Appendix 1 (Subject Excel Data) 
 
Table A1.1  Subject Demographics 
Subject Age Height (in) Weight (lb) Sex 
001 43 68 205 M 
145 35 70 209 M 
343 21 70 190 M 
358 24 66 142 F 
359 26 70 180 M 
379 23 65 135 F 
401 25 64 110 F 
405 21 69 175 M 
419 21 68 148 M 
420 21 67 165 M 
 
Average 26 67.7 165.9  
St. Dev 7.3 2.2 32  
Table A1.2.  VO
2
max data 
Subject VO
2
(ml/m/kg) Ve (L/min) Vt (L) 
145 28.7 67.61 1.64 
420 34.2 51.86 1.9 
001 25.53 64.79 1.31 
419 29.27 51.85 1.28 
359 29.26 53.09 1.64 
358 22.9 34.93 1.04 
379 18.09 27.8 1.02 
343 36.42 74.85 1.86 
401 36.57 47.12 1.24 
405 30.65 61.04 1.27 
Average 29.2 53.5 1.4 
St. Dev 5.8 14.5 0.3 
 
Table A1.3.  Performance Time data 
 Condition (?C) Performance Time (min) 
Subject 001 27 24.12
28 31.22
37 15.75
55 11.75
Subject 145 27 31.75
28 29.88
37 18.58
55 6.5
80
 Condition (?C) Performance Time (min) 
Subject 358 27 25.25
28 26.42
37 17.17
55 11.92
Subject 359 27 19.7
28 23.40
37 21.20
55 20.75
Subject 379 27 12.5
28 13.53
37 14.25
55 12.56
Subject 419 27 11.43
28 10.50
37 10.11
55 10.40
Subject 420 27 15
28 12.56
37 13.75
55 13.22
Subject 405 27 10.65
28 7.63
37 8.22
55 7.63
Subject 401 27 18.67
28 18.58
37 16.12
55 15.75
Subject 343 27 11.57
28 11.32
37 17.85
55
81
Appendix 2 (Mean Facial Skin Temperature Data) 
 
Table A2.1.  Mean facial skin temperature in ?C (recorded every 30 seconds) for each 
subject at a heat index of 27?C 
Subject Number 
001 145 343 358 359 379 401 405 419 420 Average
33.50 35.21 34.47 33.16 33.93 33.61 34.83 33.82 32.70 34.32 33.95
33.57 35.04 34.53 32.74 33.29 33.74 34.81 34.04 32.80 34.11 33.87
33.54 35.06 34.61 32.68 33.18 33.74 34.75 34.06 32.80 34.12 33.85
33.52 35.07 34.72 32.66 33.05 33.73 34.72 34.06 32.82 34.14 33.85
33.40 35.11 34.89 32.59 32.98 33.73 34.71 34.06 32.92 34.15 33.85
33.36 35.13 35.08 32.52 32.91 33.72 34.72 34.07 33.01 34.33 33.88
33.39 35.15 35.23 32.49 32.94 33.72 34.73 34.18 33.14 34.50 33.95
33.46 35.22 35.39 32.62 32.92 33.78 34.74 34.30 33.31 34.72 34.05
33.65 35.26 35.51 32.37 32.90 33.82 34.79 34.35 33.55 34.98 34.12
33.88 35.28 35.57 32.28 32.85 33.93 34.83 34.38 33.72 35.17 34.19
34.17 35.28 35.69 32.28 32.83 34.01 34.84 34.47 33.82 35.29 34.27
34.47 35.30 35.69 32.28 32.83 34.06 34.88 34.54 33.93 35.34 34.33
34.78 35.30 35.92 32.39 32.82 34.21 34.89 34.64 34.03 35.39 34.44
34.99 35.22 35.89 32.47 32.76 34.30 34.92 34.76 34.14 35.44 34.49
35.18 35.17 35.94 32.61 32.73 34.49 34.96 34.87 34.23 35.49 34.57
35.32 35.14 35.96 32.71 32.70 34.63 34.94 34.86 34.35 35.49 34.61
35.44 35.10 35.97 32.53 32.69 34.88 34.97 35.03 34.44 35.53 34.66
35.52 35.18 35.97 32.54 32.65 35.05 35.00 35.14 34.43 35.52 34.70
35.57 35.25 36.02 32.53 32.67 35.19 35.05 35.21 34.50 35.53 34.75
35.60 35.39 36.04 32.59 32.70 35.34 35.08 35.23 34.56 35.56 34.81
35.62 35.44 36.09 32.66 32.75 35.43 34.92 35.27 34.60 35.63 34.84
35.63 35.51 36.13 32.56 32.70 35.45 34.75 35.33 34.70 35.69 34.85
35.66 35.53 36.16 32.54 32.77 35.60 34.61 34.70 35.83 34.82
35.69 35.58 32.58 32.85 35.74 34.52 34.69 35.93 34.70
35.72 35.67 32.64 32.83 35.84 34.46 36.00 34.74
35.76 35.75 32.54 32.88 35.90 34.31 36.07 34.74
35.79 35.83 32.54 32.96 34.30 36.11 34.59
35.86 35.95 32.65 32.95 34.30 36.12 34.64
35.91 35.97 32.64 32.98 34.41 36.21 34.69
35.92 36.03 32.70 33.02 34.36 36.26 34.71
35.96 36.03 32.80 33.09 34.30 34.43
35.87 36.08 32.72 33.27 34.20 34.43
35.87 36.10 32.82 33.31 34.13 34.44
35.97 36.11 32.88 33.43 34.13 34.50
82
Subject Number 
001 145 343 358 359 379 401 405 419 420 Average
36.02 36.19 32.98 33.27 34.10 34.51
36.01 36.22 32.98 33.42 34.09 34.55
35.93 36.32 33.00 33.44 34.14 34.57
35.99 36.36 32.96 33.40 34.68
36.06 36.40 33.08 33.52 34.77
36.06 36.44 33.13 33.54 34.79
36.06 36.47 33.21 35.25
35.92 36.48 33.21 35.20
35.95 36.48 33.27 35.23
35.90 36.51 33.35 35.25
35.98 36.57 33.47 35.34
35.96 36.64 33.55 35.38
35.99 36.65 36.32
35.75 36.69 36.22
35.66 36.70 36.18
36.71 36.71
36.71 36.71
36.71 36.71
36.72 36.72
36.73 36.73
36.78 36.78
36.82 36.82
36.83 36.83
36.86 36.86
36.89 36.89
36.91 36.91
36.96 36.96
36.95 36.95
37.00 37.00
36.96 36.96
Table A2.2.  Mean facial skin temperature in ?C (recorded every 30 seconds) for each 
subject at a heat index of 28?C 
Subject Number 
001 145 343 358 359 379 401 405 419 420 Average
32.50 34.90 33.94 32.49 34.23 34.39 34.81 33.79 32.97 34.05 33.81
32.64 34.93 33.73 32.70 34.01 34.13 34.82 34.00 32.85 34.10 33.79
32.63 34.94 33.65 32.66 33.90 34.03 34.80 34.03 32.92 34.10 33.76
32.58 34.92 33.58 32.62 33.79 33.92 34.81 34.08 32.95 34.09 33.73
32.53 34.91 33.55 32.61 33.70 33.81 34.81 34.11 32.97 34.11 33.71
32.49 34.92 33.43 32.56 33.55 33.76 34.81 34.14 32.99 34.22 33.69
32.47 34.97 33.64 32.48 33.40 33.72 34.81 34.26 33.12 34.47 33.73
32.55 34.99 33.75 32.49 33.44 33.75 34.79 34.38 33.36 34.70 33.82
32.71 35.05 33.81 32.44 33.33 33.80 34.78 34.55 33.83 34.85 33.91
33.07 35.12 34.07 32.30 33.27 33.88 34.74 34.66 34.04 35.03 34.02
83
Subject Number 
001 145 343 358 359 379 401 405 419 420 Average
33.36 35.17 34.24 32.26 33.24 33.99 34.73 34.80 34.24 35.19 34.12
33.62 35.19 34.52 32.20 33.22 34.15 34.72 34.93 34.29 35.27 34.21
33.89 35.18 34.40 32.39 33.26 34.32 34.74 34.97 34.38 35.33 34.28
34.10 35.18 34.60 32.45 33.37 34.44 34.74 35.02 34.54 35.40 34.39
34.30 35.27 34.69 32.59 33.43 34.58 34.76 35.11 34.65 35.49 34.49
34.48 35.29 34.73 32.80 33.39 34.71 34.77 35.15 34.72 35.57 34.56
34.65 35.33 34.69 32.91 33.58 34.84 34.82 34.69 35.63 34.57
34.85 35.37 34.72 32.71 33.64 34.89 34.85 34.73 35.69 34.60
35.00 35.38 34.87 33.05 33.64 35.03 34.88 34.83 35.76 34.71
35.07 35.34 34.95 33.16 33.72 35.16 34.91 34.87 35.86 34.78
35.23 35.38 34.95 33.08 33.81 35.33 34.94 34.93 35.98 34.85
35.35 35.48 35.13 33.03 33.80 35.46 34.97 35.01 36.06 34.92
35.42 35.48 35.06 33.14 33.77 35.64 35.01 34.79
35.48 35.49 33.21 33.81 35.75 35.01 34.79
35.57 35.56 33.27 33.86 35.88 35.00 34.86
35.59 35.58 33.27 33.88 35.92 34.92 34.86
35.67 35.60 33.23 33.94 36.07 34.76 34.88
35.76 35.62 33.38 33.99 36.14 34.89 34.96
35.77 35.62 33.36 34.06 34.95 34.75
35.78 35.61 33.24 34.21 34.73 34.72
35.79 35.63 33.33 34.30 34.62 34.73
35.84 35.67 33.40 34.46 34.55 34.78
35.88 35.69 33.58 34.50 34.52 34.83
35.93 35.70 33.45 34.51 34.55 34.83
35.95 35.74 33.34 34.57 34.53 34.82
35.94 35.78 33.40 34.64 34.82 34.92
35.89 35.79 33.57 34.58 34.84 34.94
35.85 35.83 33.58 34.69 34.88 34.97
35.86 35.86 33.52 34.70 34.99
35.89 35.85 33.59 34.80 35.03
35.96 35.78 33.67 34.79 35.05
35.99 35.78 33.82 34.92 35.13
36.05 35.79 33.65 34.84 35.08
36.13 35.83 33.87 34.85 35.17
36.18 35.77 33.85 34.85 35.16
36.21 35.78 33.85 34.86 35.18
36.23 35.74 33.89 34.93 35.20
36.25 35.87 33.72 35.28
36.26 35.90 33.76 35.31
36.25 35.96 33.90 35.37
36.25 35.95 33.95 35.38
36.24 35.97 34.04 35.42
36.26 35.92 34.19 35.46
36.29 36.00 36.15
36.35 36.09 36.22
36.34 36.09 36.22
36.36 36.11 36.23
36.38 36.18 36.28
84
Subject Number 
001 145 343 358 359 379 401 405 419 420 Average
36.40 36.20 36.30
36.39 36.18 36.28
36.39 36.39
36.35 36.35
36.32 36.32
36.36 36.36
36.32 36.32
36.37 36.37
36.41 36.41
Table A2.3.  Mean facial skin temperature in ?C (recorded every 30 seconds) for each 
subject at a heat index of 37?C 
Subject Number 
001 145 343 358 359 379 401 405 419 420 Average
32.66 34.81 33.13 34.47 33.81 34.01 34.47 33.59 33.66 34.05 33.87
32.81 34.82 34.59 33.60 34.15 34.46 33.81 33.69 34.17 34.01
32.81 34.90 33.14 34.47 33.50 34.10 34.47 33.90 33.69 34.20 33.92
32.80 34.97 33.03 34.43 33.41 34.06 34.52 34.02 33.74 34.21 33.92
32.77 35.08 33.40 34.40 33.30 33.98 34.54 34.20 33.77 34.18 33.96
32.75 35.17 33.78 34.32 33.19 33.94 34.54 34.43 33.75 34.22 34.01
32.73 35.29 33.98 34.27 33.05 33.85 34.57 34.61 33.81 34.28 34.05
32.72 35.36 34.10 34.27 33.02 33.70 34.60 34.85 33.97 34.52 34.11
32.87 35.46 34.31 34.26 32.90 33.67 34.62 35.04 34.11 34.78 34.20
33.03 35.55 34.55 34.27 32.78 33.70 34.61 35.12 34.31 35.11 34.30
33.31 35.60 34.76 34.52 32.73 33.81 34.57 35.22 34.48 35.33 34.43
33.63 35.65 34.73 34.74 32.72 33.90 34.55 35.35 34.61 35.57 34.54
33.92 35.71 34.83 34.94 32.72 34.08 34.57 35.43 34.81 35.65 34.67
34.16 35.76 35.00 35.03 32.71 34.32 34.59 35.51 34.87 35.71 34.77
34.42 35.77 34.99 35.22 32.83 34.59 34.61 35.61 34.95 35.79 34.88
34.61 35.78 34.92 35.27 32.92 34.78 34.57 35.66 35.06 35.82 34.94
34.79 35.79 35.15 35.30 32.99 34.94 34.58 35.74 35.11 35.84 35.02
34.85 35.87 35.15 35.31 33.06 35.10 34.53 35.17 35.84 34.99
34.99 35.94 35.13 35.34 33.15 35.19 34.53 35.23 35.87 35.04
35.17 36.02 35.20 35.36 33.27 35.37 34.47 35.29 35.90 35.12
35.34 36.11 35.45 35.41 33.34 35.49 34.31 35.31 35.93 35.19
35.42 36.17 35.08 35.45 33.40 35.56 34.26 35.98 35.17
35.58 36.24 34.91 35.52 33.45 35.65 34.42 36.07 35.23
35.66 36.28 35.09 35.60 33.52 35.70 34.43 36.21 35.31
35.72 36.33 35.20 35.66 35.75 34.42 36.35 35.63
35.74 36.35 35.21 35.79 35.86 34.33 36.50 35.68
35.81 36.39 35.20 35.78 35.95 34.24 36.56 35.70
36.16 36.40 35.11 35.86 36.03 34.28 36.62 35.78
36.09 36.42 35.33 35.89 36.16 34.27 35.69
36.43 35.69 35.94 34.20 35.56
36.44 35.73 35.99 34.06 35.56
36.47 35.73 35.81 34.06 35.52
36.49 35.68 35.87 34.08 35.53
36.53 35.75 36.00 36.09
85
Subject Number 
001 145 343 358 359 379 401 405 419 420 Average
36.57 35.61 36.14 36.11
36.61 35.43 36.02
36.62 36.62
36.45 36.45
Mean facial skin temperature in ?C (recorded every 30 seconds) for each subject at a heat 
index of 55?C 
Subject Number 
001 145 343 358 359 379 401 405 419 420 Average
33.39 34.77 34.33 34.46 34.42 34.43 34.31 34.37 33.61 33.92 34.20
33.61 34.93 34.37 34.50 34.42 34.50 34.36 34.39 33.69 34.15 34.29
33.56 34.94 34.41 34.47 34.35 34.47 34.36 34.33 33.81 34.17 34.29
33.44 34.95 34.47 34.44 34.31 34.42 34.34 34.30 33.91 34.17 34.27
33.28 34.97 34.41 34.41 34.27 34.38 34.36 34.27 34.09 34.34 34.28
33.12 35.00 34.58 34.41 34.26 34.34 34.32 34.25 34.23 34.59 34.31
32.94 35.01 34.75 34.42 34.20 34.26 34.32 34.19 34.53 34.83 34.34
32.88 35.00 34.74 34.40 34.13 34.30 34.30 34.16 34.81 35.12 34.38
32.89 35.03 34.91 34.36 33.98 34.24 34.29 34.04 34.92 35.35 34.40
33.00 35.06 34.97 34.31 33.92 34.34 34.26 34.02 35.10 35.49 34.45
33.04 35.14 35.05 34.26 33.91 34.46 34.28 34.03 35.20 35.56 34.49
33.26 35.13 35.00 34.15 33.91 34.52 34.35 34.06 35.31 35.65 34.53
33.52 35.09 35.04 34.19 33.95 34.59 34.41 34.10 35.45 35.74 34.61
33.82 35.11 34.26 33.99 34.71 34.46 34.16 35.52 35.85 34.65
34.14 35.14 34.33 34.05 35.06 34.51 34.27 35.57 35.94 34.78
34.37 35.17 34.44 34.13 35.27 34.55 34.36 35.66 35.98 34.88
34.58 35.25 34.54 34.19 35.47 34.55 35.71 36.08 35.05
34.68 35.33 34.50 34.23 35.62 34.51 35.76 36.10 35.09
34.78 35.40 34.54 34.30 35.76 34.38 35.82 36.16 35.14
34.74 35.46 34.45 34.35 35.90 34.36 35.90 36.19 35.17
34.83 34.36 34.42 36.00 34.32 35.96 36.28 35.17
34.22 34.39 36.12 34.26 36.06 36.34 35.23
34.20 34.44 36.17 34.27 36.36 35.09
34.51 36.25 34.28 36.46 35.38
34.56 36.32 34.29 36.48 35.41
34.63 36.48 34.31 36.54 35.49
34.68 34.29 36.62 35.19
34.77 34.28 34.53
34.82 34.34 34.58
34.93 34.44 34.68
34.99 34.47 34.73
35.08 34.48 34.78
35.15 35.15
35.22 35.22
35.33 35.33
35.37 35.37
35.45 35.45
35.55 35.55
35.67 35.67
86
Subject Number 
001 145 343 358 359 379 401 405 419 420 Average
35.70 35.70
35.77 35.77
87
Appendix 3 (Acceptability Data) 
 
Table A3.1.  Acceptability (%) data (recorded every 2 minutes) for each subject at a heat 
index of 27?C 
Subject Number 
001 145 343 358 359 379 401 405 419 420 Average
80 100 80 50 70 80 60 60 80 60 72 
40 60 60 50 60 60 40 50 80 50 55 
50 20 60 50 60 20 40 40 70 50 46 
50 0 60 40 60 20 30 30 70 50 41 
40 0 30 40 60 10 20 20 60 40 32 
30 0 10 40 40 0 10 10 60 40 24 
20 0 20 40 30 0 0 10 60 40 22 
20 0 30 20  0   40 18 
20 0 30 10  0    12 
20 0 30 10  0    12 
20 0 30 10      17 
10 0 20 10 
10 0 20 10 
0 20 20 
0
0
0
Table A3.2.  Acceptability (%) data (recorded every 2 minutes) for each subject at a heat 
index of 28?C 
Subject Number 
001 145 343 358 359 379 401 405 419 420 Average
80 80 80 60 60 100 60 70 90 70 75 
60 60 80 50 40 60 60 60 80 60 61 
60 20 60 50 40 40 60 40 60 50 48 
60 0 50 50 40 20 40 30 60 40 39 
60 0 30 50 30 20 40 20 50 20 32 
60 0 30 50 20 0 40  50 10 29 
60 0 50 20 0 20   0 21 
50 0 40 20  0    22 
40 0 40 10  0    18 
40 0 30 10  0    16 
40 0 30 0      18 
40 0 20 0      20 
40 0 10 0      17 
30 0 10       13 
20 0 10 
88
Subject Number 
001 145 343 358 359 379 401 405 419 420 Average
20 20
20 20
20 20
Table A3.3.  Acceptability (%) data (recorded every 2 minutes) for each subject at a heat 
index of 37?C 
Subject Number 
001 145 343 358 359 379 401 405 419 420 Average
80 80 100 50 60 90 60 70 90 70 75
60 40 60 50 40 60 40 40 70 60 52
60 20 50 40 40 40 40 20 60 40 41
60 20 50 40 40 20 20 10 60 40 36
60 0 50 30 30 0 10 10 50 30 24
50 0 50 20 20 0 0  60 20 24
40 0 50 20 20 0 0   0 16
30 0 30 20 10 0 0   0 13
0 30 0 10  0    8
0 10      5
Table A3.4.  Acceptability (%) data (recorded every 2 minutes) for each subject at a heat 
index of 55?C 
Subject Number 
001 145 343 358 359 379 401 40
5
419 420 Average
70 80 100 50 60 80 60 60 70 60 69 
10 60 100 50 40 40 40 40 60 50 49 
0 0 30 50 40 20 40 20 50 40 29 
0 0 50 40 10 40 10 50 30 26 
0 40 40 0 40 0 60 20 25 
0 40 20 0 20  60 10 21 
 40 20 0 0   0 12 
 0 0 0
0 0
0 0
Table A3.5.  Acceptability (%)for each heat index (?C) condition over time (min) 
calculated from Equation 5-5. 
Time A(27) A(28) A(37) A(55) 
0 70.89 74.45 72.99 61.39 
0.5 67.29 70.71 68.19 57.37 
1.5 60.63 63.79 59.50 50.10 
2.5 54.63 57.54 51.93 43.75 
89
Time A(27) A(28) A(37) A(55) 
3.5 49.23 51.91 45.31 38.21 
4.5 44.36 46.83 39.54 33.37 
5.5 39.97 42.24 34.51 29.14 
6.5 36.01 38.10 30.12 25.45 
7.5 32.45 34.37 26.28 22.23 
8.5 29.24 31.01 22.93 19.41 
9.5 26.34 27.97 20.01 16.95 
10.5 23.74 25.23 17.47 14.80 
11.5 21.39 22.76 15.24 12.93 
12.5 19.27 20.53 13.30 11.29 
13.5 17.36 18.52 11.61 9.86 
14.5 15.65 16.71 10.13 8.61 
15.5 14.10 15.07 8.84 7.52 
16.5 12.70 13.60 7.71 6.57 
17.5 11.45 12.27 6.73 5.73 
18.5 10.31 11.06 5.87 5.01 
19.5 9.29 9.98 5.13 4.37 
20.5 8.37 9.00 4.47 3.82 
21.5 7.54 8.12 3.90 3.34 
22.5 6.80 7.33 3.41 2.91 
23.5 6.13 6.61 2.97 2.54 
24.5 5.52 5.96 2.59 2.22 
25.5 4.97 5.38 2.26 1.94 
26.5 4.48 4.85 1.98 1.69 
27.5 4.04 4.38 1.72 1.48 
28.5 3.64 3.95 1.50 1.29 
29.5 3.28 3.56 1.31 1.13 
30.5 2.95 3.21 1.15 0.99 
31.5 2.66 2.90 1.00 0.86 
32.5 2.40 2.61 0.87 0.75 
33.5 2.16 2.36 0.76 0.66 
34.5 1.95 2.13 0.66 0.57 
35.5 1.75 1.92 0.58 0.50 
36.5 1.58 1.73 0.51 0.44 
37.5 1.42 1.56 0.44 0.38 
38.5 1.28 1.41 0.39 0.33 
39.5 1.16 1.27 0.34 0.29 
40.5 1.04 1.15 0.29 0.25 
41.5 0.94 1.03 0.26 0.22 
42.5 0.85 0.93 0.22 0.19 
43.5 0.76 0.84 0.20 0.17 
44.5 0.69 0.76 0.17 0.15 
45.5 0.62 0.68 0.15 0.13 
46.5 0.56 0.62 0.13 0.11 
47.5 0.50 0.56 0.11 0.10 
90
Time A(27) A(28) A(37) A(55) 
48.5 0.45 0.50 0.10 0.09 
49.5 0.41 0.45 0.09 0.08 
50.5 0.37 0.41 0.08 0.07 
51.5 0.33 0.37 0.07 0.06 
52.5 0.30 0.33 0.06 0.05 
53.5 0.27 0.30 0.05 0.04 
54.5 0.24 0.27 0.04 0.04 
55.5 0.22 0.24 0.04 0.03 
56.5 0.20 0.22 0.03 0.03 
57.5 0.18 0.20 0.03 0.03 
58.5 0.16 0.18 0.03 0.02 
59.5 0.14 0.16 0.02 0.02 
60.5 0.13 0.15 0.02 0.02 
61.5 0.12 0.13 0.02 0.01 
62.5 0.11 0.12 0.01 0.01 
63.5 0.09 0.11 0.01 0.01 
64.5 0.09 0.10 0.01 0.01 
65.5 0.08 0.09 0.01 0.01 
66.5 0.07 0.08 0.01 0.01 
67.5 0.06 0.07 0.01 0.01 
68.5 0.06 0.06 0.01 0.01 
69.5 0.05 0.06 0.01 0.01 
70.5 0.05 0.05 0.00 0.00 
71.5 0.04 0.05 0.00 0.00 
72.5 0.04 0.04 0.00 0.00 
73.5 0.03 0.04 0.00 0.00 
74.5 0.03 0.03 0.00 0.00 
75.5 0.03 0.03 0.00 0.00 
76.5 0.02 0.03 0.00 0.00 
77.5 0.02 0.03 0.00 0.00 
78.5 0.02 0.02 0.00 0.00 
79.5 0.02 0.02 0.00 0.00 
80.5 0.02 0.02 0.00 0.00 
81.5 0.01 0.02 0.00 0.00 
82.5 0.01 0.02 0.00 0.00 
83.5 0.01 0.01 0.00 0.00 
84.5 0.01 0.01 0.00 0.00 
85.5 0.01 0.01 0.00 0.00 
86.5 0.01 0.01 0.00 0.00 
87.5 0.01 0.01 0.00 0.00 
88.5 0.01 0.01 0.00 0.00 
89.5 0.01 0.01 0.00 0.00 
90.5 0.01 0.01 0.00 0.00 
91.5 0.01 0.01 0.00 0.00 
92.5 0.00 0.01 0.00 0.00 
91
Time A(27) A(28) A(37) A(55) 
93.5 0.00 0.00 0.00 0.00 
94.5 0.00 0.00 0.00 0.00 
95.5 0.00 0.00 0.00 0.00 
96.5 0.00 0.00 0.00 0.00 
97.5 0.00 0.00 0.00 0.00 
98.5 0.00 0.00 0.00 0.00 
99.5 0.00 0.00 0.00 0.00 
 
0
10
20
30
40
50
60
70
80
0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30
Time (min)
A
c
c
e
p
t
a
b
i
l
i
t
y
(
%
)
HI=27
HI=28
HI=37
HI=55
 
Figure A3.1.  Average acceptability (%) vs. time (min) for all heat index conditions.  
Data taken from Table A3.5.   
 
Table A3.6.  Time (min) to reach acceptability levels (30%, 40%, and 50%) at various 
heat indices calculated from Equation 5-6 
Heat Index (?C) 30% 40% 50%
27 8.20 5.34 3.47
28 8.06 5.22 3.38
29 7.93 5.10 3.28
30 7.80 4.99 3.19
31 7.67 4.88 3.10
32 7.54 4.77 3.02
33 7.41 4.66 2.94
34 7.29 4.56 2.85
35 7.17 4.46 2.78
36 7.05 4.36 2.70
37 6.93 4.26 2.62
38 6.82 4.17 2.55
39 6.70 4.08 2.48
40 6.59 3.99 2.41
41 6.48 3.90 2.35
92
Heat Index (?C) 30% 40% 50%
42 6.37 3.81 2.28
43 6.27 3.73 2.22
44 6.16 3.65 2.16
45 6.06 3.57 2.10
46 5.96 3.49 2.04
47 5.86 3.41 1.98
48 5.76 3.33 1.93
49 5.67 3.26 1.88
50 5.57 3.19 1.82
51 5.48 3.12 1.77
52 5.39 3.05 1.72
53 5.30 2.98 1.68
54 5.21 2.91 1.63
55 5.12 2.85 1.59
93
Appendix 4 (Rectal Temperature Data) 
 
Table A4.1.  Rectal temperature in ?C (recorded every 30 seconds) for each subject at a 
heat index of 27?C 
Subject Number 
001 145 343 358 359 379 401 405 419 420 Average
37.16
37.62 37.7 37.4 37.4 37.31 37.21 37.49 37.29 37.69
37.423
37.17
37.63 37.7 37.4 37.4 37.31 37.24 37.52 37.34 37.7
37.44
37.19
37.64 37.7 37.4 37.4 37.31 37.26 37.54 37.36 37.72
37.451
37.21
37.64 37.7 37.4 37.4 37.31 37.28 37.56 37.38 37.74
37.462
37.26
37.65 37.7 37.4 37.4 37.32 37.31 37.6 37.4 37.76
37.483
37.26
37.66 37.7 37.4 37.4 37.33 37.33 37.62 37.42 37.78
37.498
37.3
37.67 37.8 37.5 37.4 37.36 37.36 37.66 37.45 37.82
37.526
37.32
37.68 37.8 37.5 37.5 37.38 37.39 37.69 37.47 37.84
37.548
37.36
37.7 37.8 37.5 37.5 37.4 37.43 37.74 37.5 37.88
37.578
37.39
37.71 37.8 37.5 37.5 37.44 37.48 37.78 37.54 37.89
37.611
37.43
37.72 37.9 37.5 37.5 37.46 37.51 37.83 37.57 37.92
37.637
37.45
37.75 37.9 37.6 37.6 37.49 37.56 37.87 37.6 37.93
37.662
37.48
37.78 37.9 37.6 37.6 37.52 37.57 37.92 37.63 37.96
37.695
37.51
37.81 37.9 37.6 37.6 37.55 37.61 37.97 37.66 37.98
37.726
37.54
37.82 38 37.7 37.7 37.59 37.65 38.01 37.71 38.01
37.76
37.57
37.85 38 37.7 37.7 37.62 37.68 38.07 37.76 38.02
37.792
37.59
37.88 38 37.7 37.7 37.66 37.69 38.1 37.81 38.06
37.821
37.63
37.89 38 37.7 37.7 37.69 37.7 38.15 37.84 38.11
37.849
37.63
37.91 38 37.8 37.8 37.73 37.72 38.19 37.87 38.12
37.873
37.67
37.93 38.1 37.8 37.8 37.78 37.72 38.24 37.89 38.16
37.905
37.68
37.95 38.1 37.8 37.9 37.81 37.73 38.29 37.94 38.19
37.936
37.72
37.97 38.1 37.8 37.9 37.85 37.74 38.33 37.98 38.22
37.965
37.74
37.99 38.2 37.8 37.9 37.89 37.77 38.02 38.26
37.79
38.01 37.8 37.9 37.93 37.82 38.05 38.29
37.82
38.04 37.9 38 37.97 37.84 38.32
37.84
38.06 37.9 38 37.99 37.87 38.34
37.87
38.07 37.9 38 37.91 38.37
37.89
38.08 37.9 38.1 37.93 38.39
37.9
38.1 37.9 38.1 37.96 38.41
37.92
38.12 37.9 38.1 37.98 38.43
37.92
38.15 38 38.1 37.99
37.92
38.17 38 38.1 38.02
37.92
38.21 38 38.2 38.04
37.92
38.22 38 38.2 38.07
94
Subject Number 
001 145 343 358 359 379 401 405 419 420 Average
37.92
38.24 38 38.2 38.09
37.96
38.27 38.1 38.2 38.11
37.96
38.29 38.1 38.2 38.13
37.98
38.32 38.1 38.2
37.98
38.34 38.1 38.3
37.99
38.35 38.1 38.3
38
38.37 38.2
38.01
38.38 38.2
38.01
38.39 38.2
38.02
38.44 38.2
38.04
38.44 38.2
38.12
38.46 38.2
38.31
38.47 38.3
38.64
38.5 38.3
38.74
38.51 38.3
38.54 38.3
38.57 38.3
38.57
38.59
38.62
38.63
38.66
38.68
38.69
38.73
38.74
38.76
38.79
38.82
38.84
Table A4.2.  Rectal temperature in ?C (recorded every 30 seconds) for each subject at a 
heat index of 28?C 
Subject Number 
001 145 343 358 359 379 401 405 419 420 Average
37.67 37.63 37.41 37.5 37.37 37.32 37.43 37.45 37.32 37.23
37.433
37.68 37.67 37.42 37.53 37.4 37.34 37.43 37.48 37.32 37.25
37.452
37.68 37.67
37.42
37.53 37.37 37.33 37.44 37.49 37.34 37.26
37.453
37.68 37.67 37.44 37.53 37.39 37.35 37.46 37.51 37.35 37.28
37.466
95
Subject Number 
001 145 343 358 359 379 401 405 419 420 Average
37.67 37.67 37.46 37.54 37.41 37.37 37.46 37.53 37.36 37.29
37.476
37.67 37.69 37.49 37.55 37.43 37.38 37.49 37.55 37.37 37.3
37.492
37.7 37.69 37.54 37.57 37.43 37.41 37.51 37.57 37.4 37.31
37.513
37.77 37.7 37.56 37.59 37.46 37.43 37.52 37.58 37.42 37.33
37.536
37.83 37.71 37.58 37.6 37.47 37.47 37.54 37.61 37.45 37.35
37.561
37.87 37.73 37.61 37.64 37.49 37.5 37.57 37.64 37.47 37.36
37.588
37.88 37.74 37.64 37.66 37.51 37.53 37.59 37.66 37.48 37.37
37.606
37.89 37.74 37.66 37.7 37.52 37.55 37.61 37.69 37.52 37.39
37.627
37.91 37.75 37.7 37.72 37.53 37.6 37.63 37.72 37.54 37.42
37.652
37.96 37.76 37.73 37.75 37.57 37.63 37.68 37.74 37.58 37.43
37.683
37.98 37.77 37.73 37.78 37.58 37.66 37.7 37.78 37.61 37.44
37.703
38 37.78 37.79 37.82 37.59 37.7 37.72 37.81 37.64 37.47
37.732
38.03 37.79 37.81 37.84 37.62 37.73 37.74 37.68 37.48
38.07 37.81 37.84 37.87 37.66 37.77 37.77 37.72 37.49
38.12 37.82 37.88 37.9 37.66 37.81 37.79 37.75 37.52
38.16 37.83 37.9 37.93 37.69 37.85 37.83 37.78 37.54
38.21 37.84 37.94 37.94 37.7 37.9 37.84 37.82 37.57
38.28 37.87 37.96 37.98 37.72 37.94 37.87 37.85 37.58
38.29 37.88 38 37.99 37.75 37.97 37.9 37.61
38.3 37.89 38.02 37.78 38.02 37.93 37.64
38.3 37.91 38.04 37.78 38.05 37.93 37.66
38.3 37.93 38.05 37.82 38.08 37.94 37.68
38.25 37.94 38.07 37.85 38.15 37.94
38.18 37.96 38.08 37.86 38.18 37.94
38.01 37.99 38.09 37.87 37.94
38.01 38.12 37.91 37.94
38.02 38.13 37.93 37.93
38.05 38.14 37.98
38.07 38.17 38
38.09 38.18 38.03
38.11 38.21 38.06
38.13 38.22 38.08
38.14 38.23 38.12
38.16 38.24 38.14
38.18 38.27 38.18
38.19 38.28 38.19
38.22 38.29 38.25
38.25 38.32 38.29
38.27 38.33 38.31
96
Subject Number 
001 145 343 358 359 379 401 405 419 420 Average
38.29 38.35 38.33
38.32 38.36 38.35
38.34 38.38 38.39
38.36 38.4 38.43
38.39 38.43
38.41 38.44
38.43 38.46
38.45 38.47
38.47 38.51
38.5 38.53
38.53
38.56
38.58
38.62
38.65
38.66
38.68
Table A4.3.  Rectal temperature in ?C (recorded every 30 seconds) for each subject at a 
heat index of 37?C 
Subject Number 
001 145 343 358 359 379 401 405 419 420 Average
37.56 37.55 37.63 37.55 37.38 37.24 37.2 37.6 37.18 36.4
37.329
37.58 37.58 37.63 37.66 37.43 37.34 37.2 37.61 37.24 36.5
37.38
37.64 37.59 37.63 37.67 37.44 37.24 37.3 37.61 37.25 36.5
37.381
37.72 37.6 37.66 37.67 37.46 37.26 37.3 37.62 37.27 36.5
37.399
37.78 37.61 37.67 37.68 37.46 37.28 37.3 37.62 37.27 36.5
37.413
37.79 37.63 37.69 37.69 37.48 37.29 37.3 37.63 37.29 36.5
37.427
37.81 37.64 37.76 37.73 37.49 37.31 37.3 37.64 37.33 36.5
37.453
37.84 37.66 37.77 37.75 37.5 37.36 37.4 37.66 37.32 36.5
37.473
37.87 37.67 37.81 37.78 37.53 37.38 37.4 37.66 37.35 36.5
37.499
37.9 37.69 37.86 37.79 37.55 37.42 37.5 37.68 37.38 36.6
37.528
37.93 37.71 37.89 37.82 37.57 37.45 37.5 37.69 37.41 36.6
37.553
37.96 37.73 37.92 37.83 37.58 37.49 37.5 37.71 37.43 36.6
37.577
38.01 37.76 37.98 37.84 37.6 37.51 37.6 37.73 37.5 36.6
37.611
38.06 37.78 37.99 37.85 37.64 37.54 37.6 37.74 37.52 36.6
37.636
37.81 38.02 37.87 37.65 37.57 37.7 37.77 37.56 36.7
37.82 38.04 37.88 37.69 37.63 37.7 37.78 37.59 36.7
37.84 38.07 37.91 37.71 37.63 37.7 37.8 37.62 36.7
97
Subject Number 
001 145 343 358 359 379 401 405 419 420 Average
37.87 38.09 37.93 37.73 37.66 37.7 37.67 36.7
37.88 38.12 37.94 37.77 37.67 37.8 37.71 36.7
37.9 38.14 37.96 37.79 37.71 37.8 37.75 36.7
37.93 38.17 37.97 37.82 37.75 37.8 37.78 36.7
37.96 38.19 38 37.84 37.77 37.8 36.7
37.97 38.21 38.02 37.87 37.81 37.9 36.7
37.99 38.24 38.04 37.91 37.83 37.9 36.7
38.01 38.26 38.06 37.94 37.87 37.9 36.8
38.02 38.28 38.08 37.97 37.95 37.9 36.8
38.03 38.3 38.09 37.99 37.99 38 36.8
38.04 38.32 38.12 38.03 38.03 38 36.8
38.07 38.34 38.14 38.07 38.11 38
38.09 38.37 38.16 38.09 38
38.1 38.39 38.18 38.13 38
38.12 38.41 38.19 38.16 38
38.15 38.45 38.22 38.18 38.1
38.17 38.48 38.25 38.23
38.19 38.27 38.26
38.24 38.28
38.24 38.33
38.24 38.36
38.4
38.43
38.46
38.51
38.54
TableA4.4.  Rectal temperature in ?C (recorded every 30 seconds) for each subject at a 
heat index of 55?C 
Subject Number 
001 145 343 358 359 379 401 405 419 420 Average
37.36
37.26
37.58 37.12 37.57 37.41 37.48 37.34 37.28
37.38
37.38
37.26
37.59 37.12 37.57 37.41 37.49 37.36 37.36
37.39
37.39
37.26
37.59 37.13 37.58 37.42 37.51 37.39 37.37
37.40
37.41
37.26
37.59 37.14 37.6 37.44 37.56 37.42 37.39
37.42
37.43
37.26
37.59 37.14 37.61 37.45 37.59 37.46 37.46
37.44
37.46
37.26
37.61 37.16 37.63 37.48 37.61 37.48 37.49
37.46
37.46
37.28
37.62 37.18 37.64 37.5 37.63 37.52 37.52
37.48
37.48
37.31
37.63 37.23 37.66 37.54 37.66 37.54 37.55
37.51
98
Subject Number 
001 145 343 358 359 379 401 405 419 420 Average
37.49
37.33
37.64 37.21 37.69 37.57 37.7 37.58 37.59
37.53
37.51
37.36
37.67 37.21 37.7 37.61 37.73 37.61 37.61
37.56
37.53
37.38
37.69 37.22 37.71 37.64 37.77 37.64 37.63
37.58
37.53
37.39
37.73 37.23 37.73 37.67 37.81 37.67 37.64
37.60
37.55
37.41
37.75 37.24 37.76 37.7 37.84 37.7 37.66
37.62
37.43
37.78 37.25 37.78 37.73 37.89 37.74 37.68
37.46
37.82 37.28 37.79 37.75 37.92 37.77 37.69
37.49
37.84 37.29 37.81 37.79 37.96 37.82 37.7
37.52
37.88 37.31 37.83 37.81 37.84 37.71
37.55
37.9 37.33 37.85 37.83 37.88 37.75
37.58
37.93 37.36 37.87 37.84 37.93 37.77
37.61
37.95 37.38 37.89 37.86 37.97 37.79
37.97 37.41 37.9 37.89 38 37.81
37.99 37.43 37.93 37.9 38.06 37.83
38.01 37.46 37.94 37.92 37.84
38.02 37.49 37.96 37.93 37.86
38.04 37.53 37.97 37.95 37.89
37.55 38.01 37.97 37.89
37.58 37.98 37.94
37.61 38.01
37.64 38.02
37.67 38.04
37.7 38.06
37.72 38.08
37.76
37.77
37.81
37.84
37.88
37.88
37.91
37.95
37.97
38
99
Appendix 5 (Performance Time ? Heat Index Model) 
 
Table A5.1.  Performance time (PT) values (min) for each subject at various heat indices 
(?C) 
HI PT (001) PT(145) PT(358) PT(359) PT(379) PT(419) PT(420) PT(405) PT(401) PT(343) 
27 27.23 31.73 25.76 21.51 13.35 10.83 13.81 9.12 18.34 11.19 
28 26.28 30.03 24.98 21.48 13.34 10.81 13.78 9.06 18.23 11.73 
29 25.35 28.42 24.23 21.46 13.32 10.79 13.76 8.99 18.11 12.29 
30 24.46 26.89 23.50 21.43 13.31 10.76 13.74 8.93 18.00 12.87 
31 23.60 25.45 22.79 21.41 13.29 10.74 13.71 8.86 17.88 13.48 
32 22.77 24.08 22.10 21.38 13.28 10.72 13.69 8.80 17.77 14.13 
33 21.97 22.79 21.43 21.35 13.26 10.69 13.67 8.74 17.65 14.80 
34 21.20 21.57 20.79 21.33 13.25 10.67 13.64 8.67 17.54 15.50 
35 20.45 20.41 20.16 21.30 13.23 10.65 13.62 8.61 17.43 16.24 
36 19.73 19.31 19.55 21.28 13.22 10.62 13.60 8.55 17.32 17.02 
37 19.04 18.28 18.96 21.25 13.21 10.60 13.58 8.49 17.21 17.83 
38 18.37 17.30 18.39 21.22 13.19 10.58 13.55 8.43 17.10 18.67 
39 17.72 16.37 17.83 21.20 13.18 10.56 13.53 8.37 16.99 19.56 
40 17.10 15.49 17.29 21.17 13.16 10.53 13.51 8.31 16.88 20.50 
41 16.50 14.66 16.77 21.15 13.15 10.51 13.49 8.25 16.77 21.47 
42 15.92 13.87 16.26 21.12 13.13 10.49 13.46 8.19 16.66 22.49 
43 15.36 13.13 15.77 21.09 13.12 10.46 13.44 8.13 16.56 23.56 
44 14.82 12.42 15.30 21.07 13.10 10.44 13.42 8.07 16.45 24.69 
45 14.30 11.76 14.84 21.04 13.09 10.42 13.40 8.01 16.35 25.86 
46 13.79 11.13 14.39 21.02 13.08 10.40 13.37 7.96 16.24 27.09 
47 13.31 10.53 13.95 20.99 13.06 10.38 13.35 7.90 16.14 28.38 
48 12.84 9.96 13.53 20.97 13.05 10.35 13.33 7.84 16.04 29.74 
49 12.39 9.43 13.12 20.94 13.03 10.33 13.31 7.79 15.93 31.15 
50 11.95 8.92 12.73 20.92 13.02 10.31 13.29 7.73 15.83 32.64 
51 11.53 8.44 12.34 20.89 13.01 10.29 13.26 7.68 15.73 34.19 
52 11.13 7.99 11.97 20.86 12.99 10.26 13.24 7.62 15.63 35.82 
53 10.74 7.56 11.61 20.84 12.98 10.24 13.22 7.57 15.53 37.52 
54 10.36 7.16 11.26 20.81 12.96 10.22 13.20 7.51 15.43 39.31 
55 10.00 6.77 10.92 20.79 12.95 10.20 13.17 7.46 15.33 41.18 
 
100
Appendix 6 (Survey Data) 
 
Table A6.1.  Perceived Effort and Reward Questionnaire Results 
Subject Score 
359 0.00 
405 1.00 
419 0.21 
420 0.51 
343 0.64 
379 0.21 
145 0.41 
401 0.54 
001 0.49 
358 0.21 
 
Table A6.2. Claustrophobia Questionnaire Results 
Subject Score 
359 0.00 
405 0.92 
419 0.65 
420 0.50 
343 0.58 
379 0.42 
145 0.81 
401 1.15 
001 1.04 
358 0.75 
 
Table A6.3.  Physical Activity Questionnaire Results 
Subject kcal/wk 
359 4396 
405 5108 
419 6316 
420 5928 
343 22624 
379 3262 
145 2078 
401 3878 
001 28616 
358 3768 
 
101
Table A6.4.  Respirator User Questionnaire Results.  Frequency represents the number of 
times the respirator had been worn prior to taking part in the study. 
Subject Frequency 
359 30 
405 6 
419 0 
420 0 
343 25 
379 8 
145 80 
401 17.5 
1 80 
358 50 
 
Table A6.5.  MBTI Personality Type Questionnaire Results 
 EI   SN   TF   JP 
Subject E I EI S N SN T F TF J P JP 
001 4 6 0.40 5 15 0.25 5 5 0.50 16 4 0.80 
145 9 1 0.90 6 14 0.30 5 15 0.25 11 9 0.55 
358 10 0 1.00 9 11 0.45 7 13 0.35 9 11 0.45 
359 4 6 0.40 14 6 0.70 3 17 0.15 13 7 0.65 
405 1 9 0.10 10 10 0.50 5 15 0.25 11 9 0.55 
419 6 4 0.60 15 5 0.75 15 5 0.75 16 4 0.80 
420 9 1 0.90 17 3 0.85 7 13 0.35 7 13 0.35 
343 5 5 0.50 10 10 0.50 7 3 0.70 11 9 0.55 
379 3 7 0.30 16 4 0.80 11 9 0.55 13 7 0.65 
401 3 7 0.30 12 8 0.60 2 18 0.10 12 8 0.60 
 
102
Appendix 7 (C
1
graphs) 
 
All C
1
values are from Equation 5-3. 
0
20
40
60
80
100
120
140
160
0 10 20 30 40 50 60 70 80
Ve (L/min)
C
1
Female
Male
 
Figure A7.1.  C
1
vs. V
e
from Table A1.2.  for each subject (n=10). 
 
0
20
40
60
80
100
120
140
160
0 0.2 0.4 0.6 0.8 1 1.2
PERQ
C
1
Figure A7.2.  C
1
vs. Perceived Effort and Reward Questionnaire results from Table A6.1. 
for each subject (n=10). 
 
103
0
20
40
60
80
100
120
140
160
0 0.2 0.4 0.6 0.8 1 1.2 1.4
CLQ
C
1
Figure A7.3.  C
1
vs. Claustrophobia Questionnaire results from Table A6.2. for each 
subject (n=10). 
 
0
20
40
60
80
100
120
140
160
0 5000 10000 15000 20000 25000 30000 35000
Physical Activity (kcal/wk)
C
1
Figure A7.4.  C
1
vs. level of physical activity results from Table A6.3. for each subject 
(n=10). 
 
104
0
20
40
60
80
100
120
140
160
0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90 1.00
EI
C
1
Figure A7.5.  C
1
vs. Extroversion/Introversion personality characteristic from Table 
A6.5. for each subject (n=10). 
 
0
20
40
60
80
100
120
140
160
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
SN
C
1
Figure A7.6.  C
1
vs. Sensing/Intuition personality characteristic from Table A6.5. for 
each subject (n=10). 
 
105
0
20
40
60
80
100
120
140
160
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
TF
C
1
Figure A7.7.  C
1
vs. Thinking/Feeling personality characteristic from Table A6.5. for 
each subject (n=10). 
 
0
20
40
60
80
100
120
140
160
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
JP
C
1
Figure A7.8.  C
1
vs. Judging/Perception personality characteristic from Table A6.5. for 
each subject (n=10). 
 
106
0
20
40
60
80
100
120
140
160
0 102030405060708090
# of times Respirator worn
C
1
Figure A7.9.  C
1
vs. Respirator Familiarity from Table A6.4. for each subject (n=10). 
 
107
Appendix 8 (Research Surveys) 
108
109
110
111
 
112
 
Respirator User Questionnaire 
 
Please answer all questions to the best of your ability. 
 
A.  Using your respirator. 
1. For how many years have you been wearing respirators? ________ 
2. How many times have you worn a respirator while performing manual labor? 
____________ 
3. If you have worn a respirator before for how long at a time do you wear a 
respirator? ______________ 
 
B. In this particular study, the respirator became a burden.  What were the reasons 
for terminating the test (circle all that apply)? 
 
 TOO HOT 
 TOO SWEATY 
 TOO HEAVY 
 TOO TIGHT 
 DIFFICULT SEEING 
 DIFFICULT BREATHING 
 DIFFICULT MOVING 
 DIFFICULT TO DO THE JOB 
 FELT AWKWARD OR CLUMSY 
 FELT SELF-CONSCIOUS 
 FELT ANXIOUS OR CLAUSTROPHOBIC 
 OTHER (please specify) ___________________________ 
 
C. Rate your attitude toward respirator masks: 
 
Unfavorable   Neutral  Favorable 
1 2 3 4 5
D.  Respirators and work. 
 
1. Rate how hard you think the following activities are WITHOUT wearing a 
respirator mask.  (Circle one for each activity). 
 
Very easy  Moderate  Very hard 
Running to catch a bus 1 2 3 4 5 
Machine working, welding 1 2 3 4 5 
Writing 1 2 3 4 5 
113
Walking through deep snow 1 2 3 4 5 
Climbing stairs 1 2 3 4 5 
Running fast 1 2 3 4 5 
Walking while carrying a 
heavy load 
1 2 3 4 5
Sweeping floors 1 2 3 4 5 
Washing clothes 1 2 3 4 5 
Shoveling fast 1 2 3 4 5 
2.  Answer these questions according to how you felt on average during each 
testing session. 
 
WHEN I USE MY RESPIRATOR: Totally 
disagree 
 Neutral  Totally 
agree 
The respirator does not interfere with my vision. 1 2 3 4 5
I stay cool.  I don?t sweat because of my respirator. 1 2 3 4 5
I can breathe easily.  The respirator doesn?t interfere with my 
breathing. 
1 2 3 4 5
The respirator is not heavy. 1 2 3 4 5
The respirator harness straps are comfortable. 1 2 3 4 5
I feel O.K. inside the mask. 1 2 3 4 5
Wearing the respirator does not interfere with my work. 1 2 3 4 5
114
Appendix 9 (Acceptability Scales) 
 
RPE 
6
7 very, very light 
 
8
9 very light 
 
10 
 
11  fairly light 
 
12 
 
13  somewhat hard 
 
14 
 
15  hard 
 
16 
 
17  very hard 
 
18 
 
19  very, very hard 
 
20 
 
115
BACS 
 
10  very, very comfortable 
 
8 comfortable 
 
6 fairly comfortable 
 
4 fairly uncomfortable 
 
2 uncomfortable 
 
0 very, very uncomfortable 
 
116
FT/OT 
 
1 very cold 
 
2 cold 
 
3 slightly cool 
 
4 neutral 
 
5 slightly warm 
 
6 hot  
 
7 very hot 
117
Appendix 10 (IRB 05-0245) 
118
119
120
121
122
 
1. Abstract: 
 
Respirators are worn in the workplace in order to protect the individual from toxic 
airborne substances; however, respirators are worn in only about 20-30% of the 
appropriate circumstances due to several factors.  The primary importance for the 
proposed research is to study the thermal effects due to increased temperature and 
humidity of inspired air.  An environmental chamber will be utilized for simulation of 
various ambient conditions, and a heating unit will supply various air conditions to the 
respirator.  Subjects will don a full face piece respirator and exercise at 65-70% of their 
VO2max at three previously determined conditions.   A mathematical model will be 
developed that correlates performance time with temperature and humidity to provide 
manufacturers with a useful tool for designing masks with favorable characteristics that 
are appropriate for optimal performance.   
 
2.  Subject Selection: 
 
a. Who will the Subjects be?  How will you recruit them? If you plan to advertise for 
subjects, include a copy of the advertisement. 
The subjects will be students at the University of Maryland.  They will be recruited from 
willing and able volunteers at the university, specifically, within the Department of 
Biological Resources Engineering.  The subjects will receive no course credit for their 
participation.  Twenty total subjects will be recruited via announcements at the 
conclusion of Biological Resources Engineering class periods.  The studies will be 
conducted at the University of Maryland. 
 
b. Will the subjects be selected for any specific characteristics (e.g. age, sex, race, ethnic 
origin, religion or any social or economic qualifications)? 
No; however, the subjects will be screened prior to testing when the informed consent 
document is completed to ensure that they are in good physical health.   
 
c. State why the selection will be made on the basis or base given in 2(b). 
The purpose of the study is to determine comfort level of the respirator, and not whether 
or not it corrects a specific problem.  For this reason, any and all volunteers will be 
selected.   
 
d.  How many subjects will participate in this protocol? 
15 subjects 
 
3.  Procedures Section: 
 
What precisely will be done to the subjects? Explain in detail your methods and 
procedures in terms of what will be done to subjects. How many subjects will be 
recruited? What is the total investment of time of the subjects. If subjects will complete 
surveys and/or other Instruments on more than one occasion, state this in the procedures 
123
section. If you are using a questionnaire or handout, include a copy within each set of 
application documents. If you are conducting a focus group, include a list of the questions 
for the focus group. 
The research will consist of three stages: obtain consent and orientation, perform the 
VO2max graded exercise test, and 65-70% VO2max testing under three different 
conditions.  The initial analysis suggests that a total of 20 subjects will be needed for the 
study. 
 
The first stage is to obtain consent and give a brief orientation to the subjects to describe 
what their participation requires.  The subjects may or may not be familiar with wearing a 
respirator; therefore, each subject will be fully advised as to the requirements of their 
participation.  The subjects will read and sign the informed consent document and fill out 
a brief medical history questionnaire.   This session provides the subject with a detailed 
description of their rights, and it provides the investigators with information about the 
subjects'' health and ability to partake in vigorous activity.  Any demographic or 
experimental data collected will remain confidential and correspond only to a subject 
number.   
 
The second stage of research consists of a graded exercise test (VO2max test) that will be 
used to determine the subject?s maximal aerobic capacity.  This test provides valuable 
information necessary for determining each subject?s work rate during each of the testing 
sessions, along with information regarding each subject?s critical termination values such 
as heart rate and maximum oxygen consumption.  Participants will be asked to warm-up 
and stretch for approximately ten minutes prior to the start of the test.  The mask used to 
collect gases during the test is a half face mask equipped with one-way inhalation and 
exhalation valves.  Heart rate measurements will be assessed using a standard ECG 
electrode configuration with the leads connected to a Patient Monitoring System.  In 
order to determine VO2max, the subject will partake in a graded exercise test to 
exhaustion.  The work rate will be adjusted every three minutes until the participant 
becomes fatigued, fails to display a rise in oxygen consumption in concurrence with the 
increase in work rate, or reaches a maximal heart rate. 
 
The third stage of research consists of actual subject testing in an environmental chamber.  
Each session will be conducted at 65-70% of the participant?s maximal oxygen 
consumption using a treadmill. All sessions will utilize a full face piece mask, which 
covers the entire face including the cheeks and forehead.  The subject will exercise at a 
constant work rate throughout all three sessions.  Furthermore, each of the subjects will 
dress similarly and wear the same mask throughout all sessions.  Before exercise, the 
subject will complete a five minute warm-up on the treadmill followed by five minutes of 
stretching.  The subject will don a heart rate monitor, rectal probe for monitoring core 
body temperature, and three surface temperature sensors on the face.  The sensors will be 
placed on the mid-forehead, right cheekbone, and upper lip under the left nostril.  The 
rectal probe will be inserted by the subject, who will be given lubricant to ease in 
insertion of the rectal probe.  The human monitor will explain the insertion technique, 
which will be to insert the tip gradually into the sphincter approximately the length of the 
palm of the subject?s hand (approximately 2 inches). The subject will be asked to dress in 
124
standard military fatigues and tennis shoes.  A full face piece respirator will be fitted to 
the subject for a comfortable, but snug fit.  The outlet of the respirator will be connected 
via a hose to monitor continuous expired airflow, and the inspiratory valve will be 
connected to the outlet of the humidity chamber.  The subject will begin to exercise at a 
treadmill speed and grade set at a work rate below 65-70% VO2max; the speed and grade 
will be increased for approximately 90 seconds before the speed and grade corresponding 
to 65-70% VO2max is reached.  At this work rate, the subject will be asked to exercise 
until he or she reaches volitional fatigue.  A 5 minute cool down period will follow the 
testing session. 
 
A human monitor while be present during subject testing to monitor for signs of distress.  
Heart rate, facial skin temperature, and core body temperature will be monitored; Rating 
of Perceived Exertion (RPE), and the Breathing Apparatus Comfort Scales (BACS) will 
be recorded each minute.  The RPE (6-20) and BACS (0-8) scales will be used to assess 
fatigue and comfort levels of the subject.  A high RPE value indicates that the subject has 
reached exhaustion, and a low BACS value indicates that the subject views the mask 
conditions as uncomfortable.  Both facial thermal and whole body thermal sensation 
values will be recorded.   
 
A participant is free to withdraw from this project at anytime without incurring a penalty. 
This request may be expressed to an investigator through either verbal or written 
communication. 
 
4.  Risks and Benefits: 
 
Are there any risks to the subjects? If so, what are these risks? What are the benefits? If 
there are known risks associated with the subject's participation in the research, what 
potential benefits will accrue to justify taking these risks? 
 
Possible risks to the subject during the maximum oxygen consumption test are minimal 
and including tripping, falling off of the treadmill, and  heat stress, though these risks are 
not anticipated.   
 
Possible risks to the subject include heat stress, severe exhaustion, or injury due to falling 
off of the treadmill.  A human monitor will be present throughout all testing sessions, and 
at any time during the testing, if a subject begins to show signs of extreme discomfort, or 
expresses extreme discomfort, the testing session will be terminated.  Due to the 
possibility of dehydration or heat stress, water will be available for the subject?s use at 
any time during testing.   
 
As protection against heat stroke, core body temperature will be monitored throughout 
the study. Testing will be terminated if rectal temperature reaches 103?F.  High rectal 
temperature is indicative of heat generated in the active muscles, and if this reaches the 
critical value, the person may become incapacitated.  As a result of normal exercise, core 
body temperature generally rises to about 101?F; however, heat stroke occurs at core 
body temperatures of 106?F; therefore, in order to avoid approaching symptoms of heat 
125
stroke, the cut-off temperature will be 103?F (McArdle, 1996).Core Body temperature 
will be monitored through a rectal probe connected to a computerized temperature 
monitoring system, after dressing has been completed. Insertion of rectal probes could 
cause some minor discomfort. The tips are round and will not hurt the subject. The rectal 
probes will be disposed off using biohazard disposal methods used in the laboratories.   
 
The study is not designed to help the subjects. Participants will receive a copy of their 
test results and this information may be of interest to those individuals interested in 
duplicating optimal environmental conditions. No monetary benefits will be provided. 
There are no direct benefits to the subjects for their participation in this study. 
 
The risks associated with this project are minimal and reversible with adequate rest. The 
benefits are specific to those persons responsive to heat stimulus and therefore will be of 
interest to this group. This benefit will not be transferred to other individuals; however, 
the minimal risk encountered makes this project feasible for this population. 
 
5. Confidentiality 
 
Adequate provisions must be made to protect the privacy of subjects and to maintain 
confidentiality of identifiable information. Explain how our procedures accomplish this 
objective, including such information as the means of data storage, data location and 
duration, description of persons with access to the data, and method of destroying the 
data when completed. If the research involves audiotaping, videotaping or digital 
recordings, state who will have access to the tapes or recordings, where the tapes or 
recordings will be kept, and state the final disposition of the tapes or recordings (i.e. Will 
the tapes or recordings be destroyed? If so, when will the tapes or recordings be 
destroyed?). 
The results of each subject will be stored under a number relating only to the order in 
which they were tested.  Subjects willing to be reached for further questions and 
comments will have a chance to give us their contact information.  Statistical information 
used in reports concerning this project will in no way be linked to any participants of the 
study. 
6. Information and Consent Forms 
 
State specifically what information will be provided to the subjects about the 
investigation. Is any of this information deceptive? State how the subject's informed 
consent will be obtained. Include a final draft of the consent form that you propose to use. 
Include a description of the data storage methods which will be used to ensure 
confidentiality within the consent form. 
Each subject will read and sign the consent form and questions regarding the study will 
be answered at this time.  Prior to the study, the subject will receive the attached 
information, including a description of the purpose of the study, the questionnaire, and 
consent form.  Only subjects who have signed the consent forms will be allowed to 
participate in the study. 
7. Conflict of Interests 
126
 
Describe the potential conflict of interest, including how such a conflict would affect the 
level of risk to the study participants. 
This is not required 
 
McArdle, W.D., Katch, F.I., and Katch, V.L., ?Exercise Physiology,? Williams and 
Winkins:  Baltimore, MD, 1996. 
127
Bibliography 
 
Culp, G. and A. Smith. 2001. Understanding Psychological Type to Improve Project 
Team Performance. Journal of Management in Engineering 17(1): 24-33. 
DuBois, A.B., Harb, Z.F., and Fox, S.H.  1990.  Thermal Discomfort of Respiratory 
Protective Devices.  American Industrial Hygiene Association Journal, 51(10):  550-554. 
 
Fox, S.H., and DuBois, A.B.  1993.  The Effect of Evaporative Cooling of Respiratory 
Protective Devices on Skin Temperature, Thermal Sensation, and Comfort.  American 
Industrial Hygiene Association Journal, 54(12):  705-710. 
 
Gwosdow, A.R., Nielsen, R., Berglund, L.G., DuBois, A.B., and Tremml, P.G.  1989.  
Effect of Thermal Conditions on the Acceptability of Respiratory Protective Devices on 
Humans at Rest.  American Industrial Hygiene Association Journal, 50(4):  188-195. 
 
Harber, P., Brown, C.L., and Beck, J.G.  1991.  Respirator Physiology Research:  
Answers in Search of the Question.  Journal of Occupational Medicine, 33(1):  38-43. 
 
James, R., Dukes-Dobos, F., and Smith, R.  1984.  Effects of Respirators Under 
Heat/Work Conditions.  American Industrial Hygiene Association Journal, 45(6):  399-
404. 
 
Johnson, A.T. 1991. Biomechanics and Exercise Physiology [online book].  Available 
www:  http://www.bre.umd.edu/johnbooks.html. 
 
Johnson, A.T., Koh, F.C., Scott, W.H., Mackey, K.M., Chen, K.Y.S., Rehak, T.  2005.  
Inhalation Flow Rates During Strenuous Exercise.  Journal of the International Society 
for Respiratory Protection, 22:  79-96. 
 
Johnson, A.T., Scott, W.H., Caretti, D.M.  2003.  Do Respirators Stress the 
Cardiovascular System?  Journal of the International Society for Respiratory Protection,
26-36. 
 
Johnson, A.T., Scott, W.H., Coyne, K.M., Sahota, M.S., Benjamin, M.B., Rhea, P.L. 
Martel, G.F., and Dooly, C.R.  1997.  Sweat Rate Inside a Full-Face piece Respirator.  
American Industrial Hygiene Association Journal, 58:  881-894. 
 
Johnson, A.T., Scott, W.H., Koh, F., Phelps, S.J.  2005.  Exercise Performance While 
Wearing a Tight-Fitting PAPR with Limited Flow.  Journal of Occupational and 
Environmental Hygiene, 2: 368-373. 
 
Klemm, F.K. and Hall, J.F.  1972.  Utility of Heat Stress Indices and Effect of Humidity 
and Temperature on Single Physiological Strains.  Journal of Applied Physiology, 33:  
436-440. 
128
 
Koh, F.  2004.  The Correlation Between Psychological Type and Performance Time 
While Wearing a Respirator.  Masters Thesis, University of Maryland, College Park. 
 
Laird, I.S., Goldsmith, R., Pack R.J., and Vitalis A.  2002.  The Effect on Heart Rate and 
Facial Skin Temperature of Wearing Respiratory Protection at Work.  Annals of 
Occupational Hygiene, 46(2):  143-148. 
 
Lekeux, A.T.  1988.  Effect of Environmental Temperature and Relative Humidity on 
Breathing Pattern and Heart Rate in Ponies During and After Standardized Exercise.  
Veterinary Record, 123(11):  295-299. 
 
Louhevaara, V., Tuomi, T., Korhonen, O., and Jaakkola, J.  1984.  Cardiorespiratory 
Effects of Respiratory Protective Devices During Exercise in Well-trained Men.  
European Journal of Applied Physiology, 52:  340-345. 
 
McCann, D.J., and Adams, W.C.  (1997).  Wet Bulb Globe Temperature Index and 
Performance in Competitive Distance Runners.  Medicine and Science in Sports and 
Exercise, 29(7):  955-961. 
 
McCuen, R.H.  1993.  Microcomputer Applications in Statistical Hydrology. Prentice-
Hall, Inc., Upper Saddle River, NJ. 
 
McCuen, R.H.  2003.  Modeling Hydrologic Change. CRC Press, Boca Raton, FL. 
 
McCuen, R.H.  1985.  Statistical Methods for Engineers. Prentice-Hall Inc., Upper 
Saddle River, NJ.   
 
National Weather Service 2005.  Heat Index. [online].  Available www: 
http://www.nws.noaa.gov. 
 
Nielsen, R., Gwosdow, A.R., Berglund, L.G., and DuBois, A.B. 1987.  The Effect of 
Temperatures and Humidity Levels in a Protective Mask on User Acceptability During 
Exercise.  American Industrial Hygiene Association Journal, 48(7):  639-645. 
 
Nunneley, S.A.  1989.  Heat Stress in Protective Clothing.  Scandinavian Journal of 
Work and Environmental Health, 15(1):  52-57. 
 
Ott, R.L. and Longnecker, M.  2001.  An Introduction to Statistical Methods and Data 
Analysis. Woodsworth Group.  Pacific Grove, CA.   
 
Payne, W.R., Portier, B., Fairweather, I., Zhou, S., and Snow, R.  1994. 
Thermoregulatory Response to Wearing Encapsulated Protective Clothing During 
Simulated Work in Various Thermal Environments.  American Industrial Hygiene 
Association Journal, 55(6):  529-536. 
 
129
Pulket, C., Henschel, A., Burg, W.R., Saltzman, B.E.  1980.  A Comparison of Heat 
Stress Indices in a Hot-Humid Environment.  American Industrial Hygiene Association 
Journal, 41(6):  442-448. 
 
Radomsky, A.S., Rachman, S., Thordarson, D.S., McIsaac, H.K., Teachman, B.A.  2001.  
The Claustrophobia Questionnaire.  Anxiety Disorders, 15:  287-297. 
 
Raffenbarger RS, Wing AL, Hyde RT.  1978.  Physical activity as an index of heart 
attach risk in college alumni. American Journal of Epidemiology,108:  161-175. 
 
Ramanathan, N.L. and Belding, H.S.  1973.  Physiologic Evaluation of the WBGT Index 
for Occupational Heat Stress.  American Industrial Hygiene Association Journal, 34(9):  
375-383. 
 
Roscoe, J.T.  1969.  Fundamental Research Statistics for the Behavioral Sciences. Holt 
Reinhart and Winston Inc., New York, NY. 
 
Selkirk, G.A. and McLellan, T.M.  2004.  Physical Work Limits for Toronto Firefighters 
in Warm Environments.  Journal of Occupational and Environmental Hygiene, 1(4):  
199-212. 
 
Tremblay, S., Radomsky, A.S., Dussault, D., Newth, S., & Brown Tesolin, S. 2002.  The 
Prediction of Effort and Reward Questionnaire. Poster presented at the Association for 
the Advancement of Behavior Therapy Conference, Reno, NV. 
 
Turner, N., Parker, J., and Hudnall, J.  1992.  The Effect of Dry and Humid Hot Air 
Inhalation on Expired Relative Humidity During Exercise.  American Industrial Hygiene 
Association, 53(4):  256-260. 
 
White, M.K and Hodous, T.K. 1987.  Reduced Work Tolerance Associated with Wearing 
Protective Clothing and Respirators.  American Industrial Hygiene Association Journal,
48(4): 304-310. 
 
White, M.K., Hodous, T.K., and Vercruyssen, M.  1991.  Effects of Thermal 
Environment and Chemical Protective Clothing on Work Tolerance, Physiological 
Responses, and Subjective Ratings.  Ergonomics, 34(4):  445-457. 
 
White, M.K., Vercruyssen, M., and Hodous, T.K.  1989.  Work Tolerance and Subjective 
Responses to Wearing Protective Clothing and Respirators During Physical Work. 
Ergonomics, 32(9):  1111-1123. 
 
Wilmore, J.H. and Costill, D.L. 2004.  Physiology of Sport and Exercise. Human 
Kinetics.  Hong Kong, China.