You are on page 1of 152

FATIGUE BEHAVIOR OF STRUCTURAL SILICONE SEALANT

by
DARREL LEWIS SHERIDAN, B.S.C.E.
A THESIS
IN
CIVIL ENGINEERING

Submitted to the Graduate Faculty


of Texas Tech University in
Partial Fulfillment of
the Requirements for
the Degree of
MASTER OF SCIENCE
IN
CIVIL ENGINEERING

Approved

Accepted

December, 1992
1 ~D~2
-?(~j_7
Jlk. ':1li3! C}3
ACKNOWLEDGEMENTS

The author wishes to acknowledge the support of the Wind Engineering


Research Center (WERC) and Glass Research and Testing Laboratory (GRTL)
of Texas Tech University. My sincere appreciation is given to Dr. H. Scott
Norville for his support and guidance. I also wish to thank Dr. Kishor
Metha for his support throughout this project. I would like to express
my appreciation to Dr. Jimmy Smith for his encouragement and support
throughout my education at Texas Tech University.
Financial support was provided by the Cooperative Program in Wind
Engineering (CPWE) by the National Science Foundation (NSF). The funds
were provided through a cooperative agreement BCS8821163. I wish to thank
those responsible for making this financial support possible.
I would like to extend a special thanks to C. B. Wright who allowed
me to work in his laboratory and offered a great deal of information,
experience, and patience. Without his involvement, this project may have
never found the next individual who is responsible for the success of this
project.
I have a most sincere appreciation for Dr. Bill Jones who designed,
built, organized, and repaired the equipment necessary for this research.
This project was able to succeed because of his involvement, expertise,
and leadership. I appreciate the time, effort, information, and advice
contributed to my education by this individual.

ii
I would also like to thank the laboratory technicians, Bill Perkins,
Frank Wyatt, and Mike McKenzie, for their involvement with this program.
They were very supportive and cooperative with me.
I would especially like to thank my parents and parents-in-law for

believing in me and encouraging me during this time. Without the support


and encouragement of my family and friends, I would not have endured to
the end.
Finally, I wish to thank my wife, Tralisa, for her love, support,
encouragement, and patience during my graduate work. She has been my
biggest supporter and best friend through it all.

iii
TABLE OF CONTENTS

ACKNOWLEDGEMENTS . . . •. . . . . . . •. . . . . . . . . . . •. . •. . . ••. . . . . ••. . . . . •. •. . . ii

LIST OF TABLES .................................................... vii


LIST OF FIGURES ................................................... ix

CHAPTER
I. INTRODUCTION ............................................. 1

Background .............................................. 1

Research Objectives ..................................... 2

Outline of Presentation ................................. 3


I I. LITERATURE REVIEW . . . • . . . • . . . . . . . . . . . . . . . . • . . . . . . . . . . . . . . . 4•

Introduction 4

Sealant Types 5

Low Modulus Sea 1ants .................................. . 5

Medium Modulus Sealants .................•...•......... 6

High Modulus Sealants 6

Silicone Sealant History 7

Introduction to Cyclic Testing .......................... 9

Review of Previous Cyclic Testing ....................... 10

Weathering and Environmental Effects .................... 12

Effects of Specimen Length .............................. 18

Goals of Testing ........................................ 20

iv
I I I. TEST PROGRAM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
Approach ................................................ 21

Research Plan ........................................... 24

Specimen Description ................................... . 24


Silicone Sealant Cyclic Testing Machine ................. 25
Introduction .......................................... 25
Test Frame ............................................ 25

Closed-Loop Feedback .................................. 30


Hydrau l i c Power Unit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
Data Acquisition ...................................... 41

Test Setup Procedure .................................... 44


IV. TEST RESULTS ............................................. 50

Introduction ............................................ 50
The Twenty psi Tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . • . . 50
The Twenty-eight psi Tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
The Thirty psi Tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
The Forty psi Tests ..................................... 67
Defective Specimens ..................................... 74
Lilliefors Test for Normality ........................... 75
Results of the Lilliefors Test for Normality ............ 77
Previous Work ........................................... 84

Regression Analysis ..•.................................. 87

v
V. CONCLUSIONS .............................................. 94

Su11111ary ••. . . . . . . . •. . . . . •. . . . •. . . . . •. . . ••. •. . •. . . . . . . •. •• 94

Conclusions . . . . . . •. . . •. . . . . . . . . •. . •. . •. . . •. . . •. . . •••. . . • 94

Reco11111endat ions . . •. . . . . . . . . . . . . •. . . . . . . . . ••. . . . . . . . ••. . • 95

Future Testing .......................................... 97

LIST OF REFERENCES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98
APPENDICES
A. SPECIFICATIONS FOR THE SSCTM PARTS ................. 100
B. LABTECH NOTEBOOK SETUP .............................. 106
C. FATIGUE DATA FOR SSG TESTING ........................ 116

vi
LIST OF TABLES

3.1 Failure Stress Levels ....................................... 23


3.2 Oscilloscope Readings for Selected Stress Levels ............ 49
4.1 SSG Test Results ............................................ 51

4.2 Quantities of the Lilliefors Test Static for Nomality ....... 78


4.3 Results of Lilliefors Test for Normality .................... 79
4.4 Results from Sandberg and Rintala ........................... 88
4.5 Results from SSG Data ....................................... 90
C• 1 SSG- 6 Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117

C.2 SSG-7 Data .................................................. 118


C.3 SSG-8 Data .................................................. 119
C.4 SSG-9 Data .................................................. 120
C.5 SSG-10 Data ................................................. 121

C.6 SSG-11 Data 122


C.7 SSG-12 Data 123
c.8 SSG-14 Data ................................................. 124
C.9 SSG-16 Data ................................................. 125
C.10 SSG-17 Data ................................................. 126
C.11 SSG-18 Data ................................................. 128
C.12 SSG-19 Data ................................................. 129
C.13 SSG-21 Data ................................................. 130
C.14 SSG-22 Data ................................................. 132

vii
c.15 SSG-23 Data ................................................. 133
C.16 SSG-24 Data ................................................. 135
c.17 SSG-26 Data ................................................. 137
C.18 SSG-27 Data ................................................. 139

viii
LIST OF FIGURES

2.1 Stress-Strain Curves for Typical Sealants .................. . 8

2.2 Full Reversal Sinusoidal Fatigue Cycle ..................... . 11

2.3 Fatigue Data for Neutral Cure Silicone Sealant 13


2.4 Separation of Window Glass from Setting Block 18
2.5 Specimen Dimension Definitions .............................. 20
3.1 Components of Structural Silicone Sealant Specimen .......... 22
3.2 Finished Structural Silicone Sealant Specimen ............... 26
3.3 Schematic of the SSCTM ...................................... 27
3.4 Components of the Hydraulic Test Cylinder ................... 28
3.5 Trans-Tek Displacement Transducer (LVDT) .................... 29
3.6 Completed Test Frame with Displacement Transducer ........•.. 31
3.7 Omega Engineering LCF-50 Load Cell .......................... 33
3.8 Moog Series 760 Servovalve .................................. 34
3.9 Flapper Nozzle Design for the Servovalve .................... 36
3.10 Moog Servocontroller Board .................................. 37
3.11 Moog Signal Conditioner Board ............................... 37
3.12 Hydraulic Power Unit for the SSCTM .....•.................... 39
3.13 Hydraulic Oil Cooler and Fan ................................ 41
3.14 Data Acquisition System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
3.15 Labtech Notebook Setup ±40 psi Test ......................... 43
3.16 Adjustable Resistor on the Signal Conditioner Board ......... 45

ix
3.17 Calibration Control on the Function Generator ............... 46
3.18 Amplitude and Frequency Controls on Function Generator ...... 48
3.19 The Oscilloscope Showing the Loading for ±30 psi Test ....... 48
4.1 Elongation Time History for SSG-12 .......................... 53
4.2 Elongation Time History for SSG-13 ......•................... 54
4.3 Elongation Time History for SSG-26 .......................... 55
4.4 Elongation Time History for SSG-27 ....•..................... 56
4.5 Elongation Time History for SSG-14 ........................•. 57
4.6 Elongation Time History for SSG-17 .......................... 58
4.7 Elongation Time History for SSG-23 .......................... 59
4.8 Elongation Time History for SSG-24 ......•..•......•......... 60
4.9 Elongation Time History for SSG-6 ........................... 62
4.10 Elongation Time History for SSG-9 ........................... 63
4.11 Elongation Time History for SSG-10 .......•......•........... 64
4.12 Elongation Time History for SSG-21 .........•................ 65
4.13 Elongation Time History for SSG-22 .................•........ 66
4.14 Elongation Time History for SSG-7 ........................... 68
4.15 Elongation Time History for SSG-8 ........................... 69
4.16 Elongation Time History for SSG-11 .......................... 70
4.17 Elongation Time History for SSG-16 .......................... 71
4.18 Elongation Time History for SSG-18 .......................... 72
4.19 Elongation Time History for SSG-19 .......................... 73
4.20 Lilliefors Test Results for ±20 psi Tests ................... 80

X
4.21 Lilliefors Test Results for ±28 psi Tests ................... 81
4.22 Lilliefors Test Results for ±30 psi Tests ................... 82
4.23 Lilliefors Test Results for ±40 psi Tests ................... 83
4.24 S-N Curve in Semi-Log Form for Sandberg and Rintala Data .... 85
4.25 S-N Curve in Hyperbolic Form for Sandberg and Rintala Data .. 86
4.26 Graphical Comparison of SSG Data versus S and R Data ........ 89
4.27 S-N Curve in Semi-Log form for SSG Data ..................... 91
4.28 S-N Curve in Hyperbolic form for SSG Data ................... 92
A.1 Change in Rated of Flow with Pressure ....................... 104
A.2 Frequency Response Change with Pressure ..................... 105

xi
CHAPTER I
INTRODUCTION

Background
A close examination of high-rise buildings in major metropolitan areas
will disclose the vast popularity of structural silicone sealants.
Structural silicone sealants have become popular because they allow
architects the ability to easily create smooth, glassy exteriors and they
offer structura 1 capab i 1it ies necessary to hold glass 1ites or stone facade
in place.
The structural silicone sealant provides a flexible, yet sturdy
attachment to the curtainwall of the building. The structural silicone
sealant must be able to provide a water tight barrier between the outside
environment and the controlled inside environment. The structural silicone
sealant must also be able to resist cyclic loadings due to wind loading,
seismic loading, and temperature fluctuations and possibly withstand harsh
environmental conditions. These environmental conditions may include
cleaning agents, moisture, acid rain, ozone, and ultraviolet light.
To date, no catastrophic failures of structurally glazed systems have
been reported. However, a more co11111on fa i 1ure of structura 1 s i 1icone
sealant occurs when small cracks develop which allow water to penetrate
into the building environment. This problem, if left uncorrected, can
lead to more serious problems. Such cracks require repair or replacement
of the structural glazing system.

1
2

Sealing costs constitute a very small fraction of the total building


costs. Structural glazing may be overshadowed by other components of the
building, such as the structural concrete and steel because it is not as
noticeable. The importance of good sealants may be completely forgotten
until exterior walls become discolored, machinery accrues damage, or carpets
are ruined. The rising energy cost required for heat and air conditioning
also bring attention to sound sealing practice (Klosowski, 1989).
In conclusion, structural silicone sealants are an important part of
the bu i 1ding. Their significance may be overlooked initially, but a
properly designed and installed sealant will pay for itself during the
life of the building.

Research Objectives
The general objective of this research is to investigate the behavior
of structural silicone sealant in response to cyclic loading. In order
to accomp 1 ish this broad genera 1 objective, severa 1 more specific objectives
must first be completed. These specific objectives include:
(1) The design and construction of a machine capable of producing a
cyclic load at various amplitudes and frequencies. A machine was
desired to produce simple cyclic loadings, such as a sinusoidal
load wave, and also non-Gaussian cyclic loadings which simulate
actual wind events.
( 2) Call ect ing data for the development of an S-N curve for the
structural silicone sealant based upon a full reversal sinusoidal
1oad ing.
3

(3) Determining the equation(s) to describe the S-N curve for the
data collected by regression analysis.
(4) Comparison of the results to previous research.

Outline of Presentation
Chapter II presents a review of previous research concerning structural
silicone sealants. This chapter primarily reviews work concerning cyclic
loading and environmental effects on structural silicone sealants. Chapter
III discusses the development of the cyclic testing machine, the test
specimens, and test procedures. Chapter IV presents the individual test
results, regression analysis results, the S-N curve, and comparison to
previous cyclic testing. The results of this research are presented in
Chapter V.
CHAPTER II
LITERATURE REVIEW

Introduction
The use of structural glazing has grown in popularity in recent years
as the result of its inherent esthetic qualities. A major component of
structurally glazed systems is structural silicone sealants. Structural
silicone sealants have excellent flexibility, strength, and movement
capabilities which allow them to be used in structural glazing.
If a structural sealant is too flexible, the glass may actually pull
away from the curtain wall. On the other hand, if the structural sealant
is too stiff, then induced thermal stresses may cause excessive localized
stresses which could lead to glass breakage or sealant failure (Schmidt
et al., 1990). Structural silicone sealants, however, balance the flexi-
bility/stiffness requirements and provide the performance levels desired
by the architect and engineer (Schmidt et al., 1989). The popularity of
structural silicone sealants will continue to grow as architects promote
the smooth, glassy finish of structurally glazed high-rise structures.
Structural silicone sealants began their steady rise in popularity in
the early 1970's and eventually became the most popular sealants in
commercial construction. In 1959, ASTM Committee C-24 on Building Seals
and Sealants was organized. Its first symposium was not held until 1988.
This symposium lead to the publication of ASTM STP 1054 "Science and
Technology of Glazing Systems." ASTM Committee C-24 held a second symposium

4
sv
in 1990 which was followed by the publication of ASTM STP 1069,"Building
Sealants: Materials, Properties, and Performance." Additional symposiums
held in 1991 and 1992 should result in similar publications, but the use
of silicone sealants has grown much faster than the understanding of the
engineering properties of structural silicone sealants in structural glazing
applications. The remainder of this chapter is devoted to reviewing briefly
previous research pertaining to the engineering properties of structural
silicone sealants.

Sealant Types
Low Modulus Sealants
Low modulus sealants typically consist of oil-based, resinous caulks,
bituminous-based mastics, and polybutene sealants (Klosowski, 1989}. The
majority of these sealants are highly filled with inexpensive fillers
(Klosowski, 1989}. Some low modulus sealants contain silicone, but these
products contain only 0.2% to 2% of a silicone fluid. The silicone fluid
is added for greater plasticity or better adhesion in most cases, but the
performance of the product is changed very little with the addition of the
s i 1 icone fluid.
The low modulus sealants are still somewhat popular as an over-the-
counter product due to their low cost. Commercial construction seldom
uses low modulus sealants due to the expense and frequency of recaulking
required. Recaulking may be required at two-year intervals.
6

Medium Modulus Sealants


This class of sealants consists of butyls, latex acrylics, Neoprenes,
and solvent-release acrylics (Klosowski, 1989). Some medium modulus
sealants are marketed as siliconized products which simply mean a small
amount of silicone has been injected into the product. The addition of
a small percentage of silicone is primarily a marketing ploy and has little
or no effect on the key performance properties of the original product
(Klosowski, 1989).
Movement capab i 1it i es are genera 11 y ±12. 5 % for med i urn modu 1us sea 1ants
which is an important advantage over low modulus sealants. The primary
advantage of using medium modulus sealants over the low modulus sealants
is an increase in service life. These products claim to have 10 to 15
years of in-service 1 ife under ideal conditions; however, the actual service
life is 3 to 10 years (Klosowski, 1989).
Medium modulus sealants maintain an advantage over low modulus sealants
because of their additional movement capabilities and their higher strength.
These two characteristics allow them to be used in many more applications.

High Modulus Sealants


This c1ass of sea 1ants consists of po 1ysul fides, urethanes, and
silicones. The most important property of these sealants is the higher
movement capabilities. Some products claim up to ±50% movement capabilities;
others provide +100% I -50% (Klosowski, 1989).
7

Sales of the high range sealants are increasing in both the over-
the-counter market and the commercial industry (Klosowski, 1989). These
products are particularly popular in the commercial industry because of
their use as a structural sealant in structural glazing. The service life
for many of the high modulus products is 20 years which is longer than the
low and medium modulus sealants.

Silicone Sealant Historv


During the early 1950's, polysulfide sealants were introduced to the
construction industry for the new curtainwall construction (Panek, 1990).
Structural silicone sealants were introduced in the early sixties (Panek,
1990). Structural silicone sealants did not gain immediate popularity
because they adhered poorly to masonry and they were more expensive than
urethane and polysulfide sealants. From 1970-1979, modifications to the
behavior of structural silicone sealants included improved movement
capabilities, better weather resistance, better adhesion, and lower costs.
These improvements made silicone sealants much more competitive with
polysulfide and urethane sealants. By 1980, structural silicone sealants
became the most popular sealants for commercial construction applications
(Panek, 1990).
Structural silicone sealants are the only sealant to have virtually
a 100% recovery which allows them to be used in structural glazing
applications (Panek, 1990). The near 100% recovery of the structural
silicone sea 1ant means that it has the ability to reg a in most of its
original physical strength and elongation abilities after a load is applied
8

and removed. This ability accounts for a large portion of the growing use
of structura 1 silicone sea 1ants. Structura 1 s i 1icone sea 1ants used in
structural glazing are exposed to cyclic loadings resulting from wind
loadings and thermal fluctuations.
Since structural silicone sealants experience cyclic loadings when
used in structural glazing, investigation is necessary to understand the
effects on their physical properties. Specifications for structural
silicone sealants do not normally include physical properties such as the
modulus of elasticity, stress-strain curves, and stress versus number of
cycles to failure (S-N curves); however, general curves or values are
available such as the stress-strain curve shown in Figure 2.1. Thus,
cyclic testing is needed to examine the design life of structural silicone
sealants by developing S-N curves.

STRESS psi

200
£,---a 7
SEALANT 'h" X 'h" X 2"
180
160 HIGH MODULUS MEDIUM
140
MODULUS
120
100
MEDIUM MODULUS
80
60
VERY LOW
40
MODULUS
20~~----------------------------------------
~~~~----~----~------------~--------~~------------------------~~
25 50 75 100 150 200 300 1000

/o ELONGATION
0

Figure 2.1: Stress-Strain Curves for Typical Sealants.


(Klosowski, 1989)
9

Introduction to Cyclic Testing


Cyclic testing is conducted to simulate the loading conditions of a
material during its in-service life span at an accelerated rate. Cyclic
testing estimates the 1ife expectancy of a material by recording the number
of cycles to failure under prescribed loads and conditions. The number
of cycles to failure is the most important information obtained from the
cyclic testing. Other data which may be recorded include deflect ion
measurements and total deflection at failure.
The number of cycles to failure is affected by the mean stress and
the maximum stress placed on the specimen. As the magnitude of the maximum
stress increases, the number of cycles to failure decreases. S-N curves
for materia 1s are generally reported for specimens subjected to full
reversal loadings. Mean stresses other than zero have an effect on the
fatigue strength of the material (Boresi and Sidebottom, 1985). Several
equations have been developed to explain the effects of the mean stress.
The Soderberg, Gerber, and Goodman relations are three descriptions of the
effects of a nonzero mean stress. The Soderberg relation approximates
conservative critical stress amplitudes for most metals, while the Gerber
relation provides good estimates of the crit ica 1 stress amplitude for
ductile materials. The Goodman relation is best suited for brittle
materials.
The primary concern of conventional endurance (cyclic) testing is the
testing of specimens under uniaxial tension-compression or rotating-bending
cycling. The results of cyclic tests usually yield plots of stress versus
number of cycles to failure (S-N curves). Many materials, such as mild
10

carbon steel, have a property known as an endurance limit. The endurance


1 imit is a stress level below which applied cyclic stresses have no apparent
effects on the material. Some materials do not seem to have an endurance
1 imit, but instead the stress continues to decrease as the number of cycles
increases. Normally for these materials, an endurance limit stress is
assigned as the stress at a given number of cycles (Boresi and Sidebottom,
1985).

Review of Previous Cyclic Testing


In applying a cyclic loading to structural silicone sealants, several
decisions must be made concerning the condition of the applied loading and
the set-up of the test. The first decision is to determine whether the
load should be a constant strain or a constant load amplitude cycling.
The structural glazing of a building is subjected to stresses caused by
wind 1oads. In previous work (Sandberg and Albers, 1980), testing of
structural silicone joints has shown that strain controlled cycles may
inhibit failure propagation. Therefore, a constant load amplitude appears
to be more appropriate.
In previous cyclic testing of 14 silicone sealant specimens (Sandberg
and Rintala, 1990), a constant load amplitude was used along with a zero-mean
stress (full reversal) cycle shown in Figure 2.2. The full reversal cycle
was chosen based on the ASTM Test for Adhesion and Cohesion of Elastomeric
Joint Sealants Under Cyclic Movement (C-719-86) as a worst case condition
(Sandberg and Rintala, 1990).
11
Full Reversal Load
20psi
30 r-----------------------------------~

20

-·-
c:ll
0..
'-"
c:ll
10

-
c:ll
II)
....
00
10

-20

-30 ~------------~------------~----------~
0 0.5 1 1.5
Time sec)

Figure 2.2: Full Reversal Sinusoidal Fatigue Cycle.

Another important factor which must be considered when testing


structural silicone sealants is the loading rate. Ideally, a fast cycling
rate is desired in order to provide quick, timely testing; however, testing
at 5 Hertz (Hz) has shown a rise in temperature within the silicone sealant
(Sandberg and Rintala, 1990; Sandberg and Albers, 1980). The temperature
change may adversely effect the results of testing; therefore, a slower
cyclic rate must be used which usually means long testing periods.
Completion of testing by Sandberg and Rintala (1980) revealed that
current design using silicone sealants in structural glazing applications
is adequate. The silicone sealants appear to be sufficient to withstand
12

the cyclic loading caused by wind loads. Sandberg and Rintala's S-N
curves are shown in Figure 2.3. However, testing should be done to study
the combined effects of environmental conditions coupled with cyclic
loadings and the effects of a non-zero mean stress cyclic load. Fatigue
and creep rupture should be studied carefully as silicones expand to new
applications and refined products.

Weathering and Environmental Effects


Structural glazing allows the architect to finish the building exterior
with a smooth finish. Structural glazing accounts for 25 to 40 percent
of new glazing (Swanson, 1987; Bailey et al., 1990). Structural silicone
sealants are used as structural elements in the transfer of external loads
from the glass to the structural frame of the building. The structural
silicone seal also functions as a barrier between the interior of the
building and the outside environment. Thus, a failure has occurred either
when water penetrates the seal or when the seal no longer transfers the
load to the structural frame. As the use of structural silicone sealants
in structural glazing continues to grow, the effects of weathering and
environmental conditions become more important.
Numerous tests have been conducted regarding temperature effects on
tens i1 e or shear strengths of s i1 icone sea 1ants. A 1imited number of tests
have been conducted relating the effects of environmental conditions, such
as acid rain, harsh detergents, cleaning agents, and ultraviolet light on
the engineering properties of silicone sealants.
13

'"'
a
CL
t. JOO
Iaw D CD
It

..,Iii
cr:: 200
:I
..1

100

0~------.--------r-------r------~------~------~
0 2

LOC(CYcUs 10 FALURE)

(a) - Semi-log form .

~
~

'"'
a
CL
.X
._,
JOO-
Iaw l
It

..,Iii
It 200-
:I
..1

~ CD 0

m IJ
a
100-

0 2D '0 EO 110 100 120


(l hou.anda)
CYCL£S TO FAUJRE

Figure 2.3: Fatigue Data for Neutral Cure Silicone Sealant.


{Sandberg and Rintala, 1990)
14

Tensile adhesion tests were performed on silicone specimens of 1/2 x


1/2 x 2 in. to determine the effects of extreme temperatures using three
different sealants {Schmidt et al., 1989). The three sealants were a high
modulus, a medium modulus, and a two-part high modulus sealant. The
experimental design placed the samples at extreme temperatures (-4~ For
175° F) for a minimum of thirty minutes. The samples were then placed in
an Instron machine and pulled at a rate of two in./min. The Instron machine
was also equipped with an environmental chamber, thus allowing the samples
to be tested under extreme temperatures that closely resembled actual field
conditions. Results of these tests revea 1ed that a11 three sea 1ants "remain
relatively constant" {Schmidt et al., 1989) at -4~ F, but exhibit some
stiffening. The stiffening is expected because all materials stiffen as
the crystallization temperature is reached and surpassed. Most organic
sealants would become brittle at -40° F because the crystallization tem-
perature is in the -20° F to oo F range.
The test at 175° F revealed a slight softening of the sealants, but

this shift was within one standard deviation of the mean of room temperature
data in all cases. Structural silicone sealants appear to be the only
sealant with the stability required for structural glazing applications
at extreme temperatures {Schmidt et al., 1989).
Another experimental program {locket al., 1990) investigating tem-
perature and environmental effects on silicone sealants revealed similar
results. That program used high modulus and medium modulus specimens cut
from cured sheets of the product. The test procedure again used an Instron
15 ~

machine to provide uniaxial tension. Various changes in test conditions


for the samples included temperature, strain rate, moisture, pH, dissolved
oxidants, and time of testing. The concept of an ideal elastomer, which
was previously shown to describe the stress-strain relationship of silicone
sealants below fifty percent elongation (Tock et al., 1987), was used to
reduce the data.
The results of Tack's tests provide interesting results. One year of
aging appeared to slightly increase the crosslink density, thus signalling
a gradual embrittlement of the silicone. Both sealants were slightly
plasticized by moisture. The dissolved oxidants had virtually no effect
on the samples. Acidic immersion produced very different results from the
two different silicone sealants. The high modulus samples when immersed
in low pH solutions appeared to experience surface aging, while the medium
modulus samples became very soft and appeared to lose their structural
capabilities. The medium modulus silicone sealant reacted violently with
the acid. Tock believed this reaction was the acid attacking the carbonate
fillers within the silicone; however, the use of such fillers was not
confirmed with the manufacturer. In both cases, exposure to acidic solutions
seemed to cause accelerated aging and structural degradation. Although,
the high modulus sealant immersed in acidic solution reached about the
same stress level, the elongation was approximately one-tenth of the samples
not immersed in the acidic solution.
Temperature had virtually no effect at low levels, but did reduce the
strength of samples at temperatures of 212° F. The reduction of strength
was not significant however. The tests reveal that the samples immersed
16

in basic solutions were not affected significantly. The samples placed


in acid solutions of pH at or below 2.0 were affected seriously. A pH of
2.0 is much lower than expected during precipitation, but may be experienced
if pooling occurs on exposed silicone followed by the evaporation of water
(Melander, 1989; lock, 1990).
In 1981, a research program began which was designed to investigate
many factors concerning structural silicone sealants (Schmidt et al.,
1989). This research included full-scale testing as well as laboratory
specimens for testing. Testing environments included controlled conditions
in the laboratory and field conditions outdoors with little to no control
over the environment. The full scale testing was conducted to determine
the structural capabilities of sealants based on their modulus values.
Also included within the study of the full scale specimens were the effects
of two years of weathering in Texas and the use of clear versus reflective
glass panels.
Thirty-one full-scale panels were constructed for testing. Sixteen
of these panels were tested immediately after adequate curing, while the
remaining fifteen panels were placed in a southern exposure for two years.
Each panel was glazed using a four-sided structural glazing design and the
entire dead load was supported by the silicone sealant. The panels were
designed using current industry standards with the trapezoidal load
equation. A design wind load of 60 psf and a structural silicone sealant
design strength of 20 psi were used to determine the proper glue line and
bite dimensions.
17

The test panels were fastened to a pressure chamber and subjected to


a simulated negative wind load resulting in tensile stresses acting on the
structural silicone sealant. Deflections were measured at the quarter
points along all edges using strain gages. Deflection readings were taken
at loads of 30, 60, and 90 psf. Each panel was loaded to failure after
the final deflection was recorded at 90 psf.
The results of this testing revealed that structural properties of
silicone are not dependent upon whether the glass surface finish was clear
or reflective. The high modulus silicone sealant became slightly less
flexible after two years of weathering. The deflection data for the medium
modulus silicone sealant was consistent for both factors considered in
the testing. The deflections for the medium modulus sealant were statis-
tically the same as the high modulus sealant. Both the medium and high
modulus silicone sealants behave in similar fashions within the design
stress of 20 psi (Schmidt et al., 1989).
The low modulus silicone sealant behaved similarly to the two higher
modulus sealants at loads of 30 psf. However, at the loads of 60 and 90
psf, the low modulus sealant deflected excessively. large deflections are
undesirable and could be dangerous if they allow the glass or other glazing
to fall from the setting block as shown in Figure 2.4.
Schmidt's research concluded that low modulus sealants should not be
used in structural applications because of their poor performance after
two years of aging. The poor performance of the low modulus sealant was
due to the loss of adhesion to reflective glass. The medium and high
modulus silicone sealants were found to be acceptable in structural glazing
18

applications. Neither of these two sealants showed any significant effects


from weathering for two years. The ultimate failure strength of the two
sealants was well above the glass design load of 60 psf and in many cases,
the glass failed before the silicone sealant (Schmidt et al., 1989).

Effects of Specimen length


Proper design of structura 1 glazing using structura 1 s i 1 icone sea 1ants
requires an understanding of engineering properties. Two important
properties of structura 1 s i 1icone sea 1ants used in structura 1 glazing
applications are the strength and stiffness. Most structurally glazed
joints are very long; however, the laboratory specimens used to determine
the engineering properties are short (Sandberg et al., 1989). Therefore,
testing was conducted to determine the effects of specimen length on the
properties of silicone sealants.

Figure 2.4: Separation of Window Glass from Setting Block.


(Schimidt et al., 1989)
Silicone sealant properties are dependent on joint geometry. As the
width increases with respect to the depth, the strength and stiffness
modulus decrease (Sandberg et a1 . , 1989). Figure 2. 5 depicts specimen
dimension definitions. Overestimation of strength in laboratory specimens
would lead to smaller safety margins in glazing design. Stiffness or
movement capabilities based on laboratory specimens need to be reliable
because overestimation or underestimation of this value could produce
problems. If the stiffness were higher than the assumed value, then
excessive stresses may lead to glass or sealant failure. If the stiffness
was lower than the assumed value, then excessive deflections could occur.
Testing was conducted on specimen lengths of 2, 4, 8, and 16 inches. The
width and depth for all specimens were one-quarter and one-half inch,
respectively. Both shear and tension loads were investigated.
The results of tensile and shear strength indicated that the specimen
length has no effect statistically. Stiffness of the silicone elastomeric
sealant for the shear and tension test again found the specimen length had
no effect statistically (Sandberg et al., 1989). This study also found
that sensitivity of workmanship flaws to specimen length was not significant.
Therefore, the laboratory specimens provide an adequate prediction of the
engineering properties used in structural glazing design.
'· I
20

..j
EPTH 14----L EN GT H _..,.

Figure 2.5: Specimen Dimension Definitions.


{Sandberg et al., 1989)

Goals of Testing
The ultimate goal of this research is to develop an S-N curve for GE
Ultraglaze SSG-4000 under full reversal zero mean stress cyclic loading.
By investigating different loading conditions, trends in the structural
silicone sealant should become apparent. Another goal of this research
is to develop a cyclic testing device which is capable of responding to
various load functions developed through research at Colorado State Uni-
versity.
The results of this research are compared to previous work by Sandberg
and Rintala {1990). Comparison with earlier work will provide validation
for this research and possibly disclose improvements in structural silicone
sealants.
CHAPTER III
TEST PROGRAM

Approach
Structural silicone sealant is subjected to cyclic loadings, such as
wind and thermal fluctuations, as well as environmental factors such as
moisture, acid rain, detergents, ozone, and ultraviolet rays. Fatigue
testing is necessary to determine how well the structural silicone sealants
perform under cyclic loading. Tests may include environmental effects in
order to determine the effects of each environmental factor. As a result
of the cyclic testing, the properties of structural silicone sealants
should be better understood and possibly improved. Such improvements can
include increased design strengths for example.
The data reported was obtained by controlled cyclic testing of 22
structurally glazed silicone sealant specimens. The specimens were provided
through joint cooperation between General Electric (GE) Silicones and
Curtain Wall Design and Consulting (CDC). Funding was provided under the
Cooperative Program in Wind Engineering (CPWE) at Texas Tech University
(TTU) through the National Science Foundation. Each specimen consisted
of a 6 x 2 x 2 in. aluminum block, 6 x 2 x 1/2 in. piece of annealed glass,
and a 2 x 1/2 x 1/2 in. strip of GE Ultraglaze SSG 4000 structural silicone
sealant. The glass was structurally glazed to the aluminum block by
structural glazing professionals to simulate curtainwall construction.
The components of the specimen are shown in Figure 3.1.

21
22

The 22 structurally glazed specimens were tested under axial loading.


The first tests were conducted with a zero mean stress (full revers a1
cycle) at 20, 28, 30, and 40 psi stress amplitudes. The failure stress
levels and number of specimens tested at each level are shown in Table
3.1. The pertinent data obtained included the number of cycles to failure,
visual observations, the number of cycles to achieve fifty percent elon-
gation, and a time history of the deflection data. Fifty percent elongation
is the working range for this structural silicone sealant. The visual
observations include the location and progression of the tear leading to
failure. The experimental design and procedures followed to conduct the
tests are described in this chapter.

Figure 3.1: Components of the Structural Silicone Sealant Specimen.


23

Table 3.1
Failure Stress Levels

Variable Number of
Stress Specimens
(psi)
20 5
28 4
30 5
40 7
24

Research Plan
The structural silicone sealant specimens were divided into four groups
comprised of four to five specimens per group. The number of cycles to
failure were determined and S-N curves (stress versus number of cycles)
were developed for a full reversal stress cycle. A comparison with previous
cyclic testing by Sandberg and Rintala (1990) should determine if the
results are complementary to one another. The data obtained provides
values to be compared to previous, as well as future testing programs.

Specimen Description
The test specimens were prepared by CDC. The aluminum block was
purchased as a long slender bar and cut to proper lengths. The annealed
glass pieces were purchased from TempGlass Southern, Inc. The silicone
sealant used for this research program was General Electric Ultraglaze SSG
4000 which is one of the strongest one-part neutral cure sealants.
Each of these three components were sent to CDC to be structurally
glazed. Each specimen was professionally glazed following industry
standards and specific product specifications. The aluminum blocks were
cleaned and treated using methyl ethyl keytone (MEK) to insure proper
adhesion. MEK is one of the best cleaning agents available for structural
glazing applications; however, MEK is seldom used today because of the
strict handling and disposal regulations set forth by the Occupational
Safety and Health Administration. Since the specimens were glazed in a
laboratory by trained professionals, MEK was used. MEK cleaning should
result in a clean surface and a strong bond between the silicone and the
25

aluminum block. The silicone was tooled and formed using Teflon block
forms provided be GE Silicones. Once the specimens were formed, they were
clamped together in order to insure full bonding between the components.
The specimens were allowed adequate curing time before being returned to
the laboratory for testing. GE reports that GE Ultraglaze SSG 4000 will
achieve full cure in 21 days. Finished test specimens are shown in Figure
3.2.

Silicone Sealant Cyclic Testing Machine


Introduction
Since cyclic testing of this nature had not been conducted at the
Glass Research and Testing Laboratory (GRTL) previously, a test frame and
a cyclic loading machine had to be built. The four components of the
testing device are the test frame, the closed loop feedback, the hydraulic
power unit, and the data acquisition system. Each component consists of
several parts which will be explained within this chapter. A schematic
of the Silicone Sealant Cyclic Testing Machine (SSCTM) is shown in Figure
3.3.

Test Frame
The test frame was built by Civil Engineering laboratory technicians
at TTU and is constructed of 1/2 structural steel plate.
11
Housed within
the test frame are the test cylinder and the Linear Variable Displacement
Transducer (LVDT).
26

The hydraulic test cylinder used for this testing is a Nopak Class 7
Square Barrel Hydraulic Cylinder, CL7-l x 6-C-AA-.375-[.375-24], from
Galland Henning Nopak, Inc. This hydraulic cylinder is constructed of
seamless extruded aluminum alloy, drawn to a 16 micro inch finish. In
order to insure a durable surface, the cylinder bore is hard coated. Life
tests have shown no appreciable wear in the bore after millions of cycles.
The square barrel provides an exclusive opportunity for mounting into any
side or either end. The cylinder heads and head plate are precision cast
and machined aluminum alloy. The parts of the hydraulic test cylinder are
shown in Figure 3.4.

Figure 3.2: Finished Structural Silicone Sealant Specimen.


27

DATA HYDRAULIC CLOSED


ACQUISITION POWER LOOP
SYSTEM UNIT FEEDBACK

r:l
DISPLACEMENT

LVDT
1:
CYLINDER ~~
SPECI-
MAN § LL
J.. ..J (ACTUATOR) ~v _J
0 J FORCE
_J
w ::J
1-
..J
.).

"
ii T T 0 0..
,. 0
a
~
I"L..-....1'
'-a::<~o ~ :1
DATA
tJ RETURN tvtoOG
&,Q ~~
~~ 1...(
t.. ..........-
SIGNAL.
w
0 1-
0 0..
::J

gz
w
SERVO jCONDITIONER
ACQUISITION
VALVE a~~
~
I
_J
j\L
\ Ct a _J

T ~1-L
0 ~
z
,I,..
f'
RELIEF'
VALVE
~t-
~
~ ~0
~ CONTROLLER
SERVO r-
r--

I
~
0
w
0:
(ij
w
,~.-
HYDRAULIC 0
' POWER
SUPPLY FUNCTION
GENERATOR

Figure 3.3: Schematic of the SSCTM.


28

1. SOCKET HfAD SQEWS (4) ROD END

Figure 3.4: Components of the Hydraulic Test Cylinder.

The test cylinder is used to apply a cyclic loading to the test specimen
which is rigidly attached to the test frame. The cyclic loading applied
by the test cylinder is controlled by the flow of hydraulic fluid. The
test cylinder is attached to the outer wall of the test frame using four
bolts and is also supported by the inner wall of the test frame. This
attachment provides adequate support to keep the test cylinder stab 1e under
the cyclic movement of the piston.
The linear Variable Displacement Transducer (LVDT) used to measure
deflection information for this project was a Trans-Tek Displacement
Transducer DC-DC Series 240. The displacement transducer is housed in a
stainless steel housing for optimum protection. The coil assembly,
oscillator-demodulator, and Teflon insulated leads are enclosed in epoxy
resin for protection. The displacement transducer is shown in Figure 3.5.
29

Figure 3.5: Trans-Tek Displacement Transducer (LVDT).

The displacement transducer relies on three components to accurately


measure displacements. These three components of the transducer are a
precision linear variable differential transformer, a solid state oscil-
lator, and a phase-sensitive demodulator. These components are designed
and housed in one small package and provide outstanding linearity, sharp
resolution, and high sensitivity. The phasing and harmonic problems
encountered with AC differential transformers are eliminated.
The circuits with in the displacement transducer are arranged in series
opposition in order to produce a resultant output of DC vo 1tage. The
30

output voltage is linearly proportional to the core displacement from the


center and the voltage polarity is a function of the direction of the core
displacement. Therefore, the d i sp 1acement transducer can accurately measure
displacement of a specimen subjected to cyclic loadings.
The structural steel frame was designed to withstand loads significantly
larger than those anticipated in the scope of this particular project.
The test frame was designed with the expectation of testing larger scale
specimens. The completed test frame with the cylinder and displacement
transducer installed is shown in Figure 3.6.

Closed-loop Feedback
The SSCTM is capable of producing several different load functions at
various frequencies. The closed-loop feedback is responsible for providing
the load function and controlling the servovalve to follow the load function.
The closed-loop feedback system has several components which include a
function generator, a load cell, a servovalve, and a hydraulic power supply
system. Each of these components are discussed in this section.
The function generator is one component of the closed-loop feedback
system. The servovalve relies on the closed-loop feedback to provide the
load wave and load magnitude specified; thus, the closed-loop feedback
system serves as the brain for the servovalve. The function generator
used for the SSCTM is a model TD 6322-10 from Edlie Electronics Company.
The purpose of the function generator is to provide a dynamic signal
which will cause the cylinder to cycle between two points of load or strain.
31

Figure 3.6: Completed Test Frame with Displacement Transducer.

The dynamic signal may be in the form of a sinusoidal, square, triangle,


or saw tooth wave in either standard or inverted form. These load waves
may be in tension, compression, or alternating tension-compression loading.
The allowable frequencies range from 0.001 to 1000 Hz; however, the
maximum frequency is dependent upon individual test conditions. These
32

test conditions include type of specimen, stroke length, volume of hydraulic


fluid available, and desired load. The frequency utilized for this project
was 1 Hz.
The load cell used for the SSCTM is a product of Omega Engineering,
Inc. The 1oad ce 11 is a tens ion/compress ion 1oad ce 11 to a11 ow full
reversal cycles or variations of tension or compression cycles. The model
number of the load cell is LCF-50.
The LCF-50 load cell is very small yet capable of providing accurate
load readings. The LCF-50 load cell is designed with male thread studs
for in-line load attachment thus allowing compression or tension loads.
The LCF series of load cells uses stainless steel welded construction to
minimize or eliminate the effects of off axis loads. The LCF series of
load cells is also constructed using an internal triple stack design to
ensure excellent long term stability.
Placed directly in line with the cylinder, the load cell measures the
force exerted on the specimen and re 1ays this information to the servova 1ve.
The load cell is part of the closed-loop feedback system for the servovalve.
The load cell provides the load magnitude which the servovalve monitors
and adjusts accordingly. The LCF-50 load cell is shown in Figure 3.7 and
the manufacturer's specifications are given in Appendix A.
The servovalve used for the SSCTM is a product of Moog Controls. The
Moog 760-910 servovalve was selected for this application. The 760 Series
is a high performance, two-stage design that covers a range of rated flows
from 1 to 15 gpm at a 1000 psi valve drop. The output stage is a closed
center, four-way, sliding spool, while the pilot stage is a symmetrical
33

double-nozzle and flapper, driven by a double air gap, dry torque motor.
The mechanical feedback for the spool position is provided by a simple
cantilever spring. The valve design is simple and rugged for dependable,
long life operation. The specifications for the Moog 760 Series servovalve
are shown in Appendix A.
Moog Control two-stage servovalves are used to provide precise control
of position, velocity, and force in a wide range of applications. Some
of these applications are very complex and require small tolerances and
high reliability. Moog Control Servovalves are found on aircraft controls,
flight simulators, and oil exploration equipment just to mention a few
uses. The servovalve is shown in Figure 3.8.

Figure 3.7: Omega Engineering LCF-50 Load Cell.


34

Figure 3.8: Moog Series 760 Servovalve.

In a closed-loop feedback system, hydraulic flow is applied to a


hydraulic actuator which provides the load. The system contains the
appropriate transducer to measure the desired parameter (position, force,
velocity, etc.) which is relayed back to the valve for comparison to the
reference signal. The difference between these signals, i.e., the error
is amplified, providing a signal to the valve to drive the hydraulic
actuator to provide the desired load function.
The primary, or first, stage of many electrohydraulic servovalves is
achieved by using a nozzle flapper valve. This device is well suited for
this use because it requires only small magnetic forces, and thus, needs
35

sma 11 electrical input power to achieve a particular function (Morse,


1963). The hydraulic amplifier which implements the nozzle flapper design
is shown in Figure 3.9.
The three basic elements of the servovalve are the torque motor, the
hydraulic amplifier, and the valve spool. When properly implemented, the
system operates in the following manner. The torque motor is driven by
an electrical signal and generates a magnetic force which acts on the ends
of an armature. This action causes a deflection of the armature/flapper
assembly in the nozzle flapper design. Fluid flow is restricted through
one nozzle by the deflection of the flapper which causes the spool to be
displaced.
Spool movement opens the supply pressure port (Ps) to one control
port, while opening the return port (R) to the other control port. The
feedback spring is also controlled by the spool motion. The spool movement
causes a restoring torque on the armature/flapper assemb 1y. When the
restoring torque is equa 1 to the torque of the magnetic forces, the
armature/flapper returns to the neutral position. The spool remains open
in a state of equilibrium until the input signal changes.
The spool position is proportional to the electrical input signal.
With a constant pressure drop across the valve, the flow to load is
proportional to the spool position. Appendix A contains hydraulic char-
acteristic and flow-load characteristic information.
36

AMATURE

Figure 3.9: Flapper Nozzle Design for the Servovalve.

The servovalve is controlled by the Moog Model 121-132 Servocontroller


and the Model 123-135 Signal Conditioner. The Servocontroller has a
built-in regulated DC power supply capable of driving two or three ser-
voamplifiers or other devices. A servoamplifier is built into the Ser-
vocontroller board and drives the servovalve with a DC current having high
dynamic response and negligible drift. The signal conditioner provides
stable amplification for low level signals such as those received from a
load cell or displacement transducer. This model accepts two input signals
which may be amplified over a wide range. This board also includes
provisions to sum the signals with precision weighing. The servocontroller
and signal conditioner are shown in Figures 3.10 and 3.11, respectively.

Hydraulic Power Unit


The hydraulic power unit consists of a hydraulic reservoir, a two HP
motor, an in-line filtration, and a hydraulic pump. Each of these products,
with the exception of the in-line filtration assembly, were acquired from
37

.. - l.;ll()

'

Figure 3.10: Moog Servocontroller Board.

ll

Figure 3.11: Moog Signal Conditioner Board.


38

Applied Energy Company, Inc., of Dallas. The components are manufactured


by different companies, but were purchased and assembled by Applied Energy
Company before being shipped to TTU. The hydraulic power unit is shown
in Figure 3.12.
The hydraulic reservoir provides storage for the hydraulic oil as well
as mounting for the motor and low pressure filtration. Also mounted on
the hydraulic reservoir is a thermometer to monitor the oil temperature.
As the oil temperature increases, the oil viscosity decreases and the
volume in the tank decreases. In order to maintain a constant pressure
inside the hydraulic reservoir, the reservoir must be able to breath in
11 11

order to replace the volume evacuated as the volume of oil decreases and
vice-versa. Thus, the hydraulic reservoir has an opening which allows air
to enter or exit the reservoir. Because the SSCTM is operated in a
laboratory environment, the air contains various impurities which pose a
threat to the servovalve. This requires the use of a filtration assembly
to remove any impurities which may be present.
Mounted on top of the hydraulic reservoir is the two HP Baldor motor
and controller box. The motor provides power to drive the pump and deliver
the required flow rate and supp 1y pressure. A simp 1e start/stop push
button switch is housed in the controller box. Also housed in the controller
box are two safety features which shut down the hydraulic system to prevent
damage to the system. One of these features is a relay switch which
monitors the temperature of the motor. If the motor overheats, this relay
switch shuts down the motor. The relay switch must be manually reset and
the motor must be given adequate time to cool before the motor may be
39

started again. The second feature is again a relay switch which is used
to power down the hydraulic system. This switch monitors the elongation
information throughout the entire test. When the elongation reaches a
desired value, the relay switch will shut down the hydraulic system to end
testing. The switch is adjustable, so the operator can select the value
which ends the testing.

Figure 3.12: Hydraulic Power Unit for the SSCTM.


40

The servovalve is a highly sensitive piece of equipment and requires


strict filtration. Particles larger than two microns will clog and possibly
damage the servovalve. Therefore, the servovalve requires a high pressure
two micron filter assembly which is placed in the hydraulic 1 ine, i11111ediately
prior to the servovalve. The high pressure filter and assembly was included
with the servovalve from Moog Controls. In order to comprehend how small
two microns is, recall the human hair is approximately forty to eighty
microns.
A second filter assembly is located on the hydraulic reservoir. This
five micron filter is a low pressure filtration system which is placed in
1ine with the return 1ine to provide filtration prior to entering the
reservoir. The primary function of this filter is to catch metal particles
and flakes from the cylinder as it wears, but this filter also provides
filtration for the bypass relief pressure line.
In addition to the in-line filtration, a hydraulic oil cooler is also
placed in the system. This hydraulic oil cooler operates in a manner
similar to that of a car radiator. The oil is forced through the cooler
which has several aluminum fins to radiate the heat from the oil and out
to the atmosphere. A small fan continually blows cool air on the hydraulic
oil cooler. The hydraulic oil cooler is shown in Figure 3.13.
The final component of the hydraulic power unit is the Webster 3081-R
pump. This unit was selected to provide a supply pressure and flow rate
compatible with the servovalve. The Webster Champion 8 Series pumps are
standard products so they are readily available and adaptable to a wide
41

Figure 3.13: Hydraulic Oil Cooler and Fan.

range of applications. The bodies of these pumps are constructed of high


density iron with heat treated alloy steel gears and shafts. The B series
also features double lip seals and heavy duty bearings.

Data Acquisition
The data acquisition system consists of a IBM 286-SX computer, an
input/output (1/0) board, and Labtech Notebook. This equipment is readily
available in most laboratory environments. The data acquisition system
is shown in Figure 3.14.
42

The computers used for the data acquisition were an IBM XT and an IBM
286. Only one computer is needed, but due to computer malfunction two
different computers were used to complete the testing. The 286 computer
obviously worked more efficiently because of the additional speed present.
However, the XT computer performed sufficiently to collect data.
The I/0 board used was a Ca.puterBoards, Inc. CIO-ADOS. This board
allows the computer to become a medium speed data acquisition system. In
addition, this board includes software which allows programmed control in
BASIC, C, FORTRAN, and PASCAL as well as many data acquisition programs
which are commercially available.

11111111111111111

Figure 3.14: Data Acquisition System.


43

The final component of the data acquisition system is a commercially


available program called Labtech Notebook. This program is menu driven
and provides the ability to collect and record data, as well as control
the system based upon the data collected. Labtech Notebook provides the
versatility to fit a wide range of applications. This program can accept
from one to fifty channels which may be input, output, or internal such
as time or calculated channels. The typical setup for the ± 40 psi test
is shown in Figure 3.15 with the setup for each channel shown in Appendix
B.

Figure 3.15: Labtech Notebook Setup for± 40 psi Tests.


44

Jest Setuo Procedure


The SSCTM is a precise instrument which provides various loadings with
a wide range of frequencies. Testing reported herein consisted of a sine
wave 1oad ing condition at one hertz (Hz} with a zero mean and various
amplitudes.
Several steps are necessary in order to begin the SSCTM machine at
the initiation of each new test. These steps are taken in order to insure
that the specimen is subjected to the desired conditions such as a load
controlled sine wave function. At the conclusion of these steps, the
machine is monitored throughout the duration of testing to insure the
proper loading is maintained and also to make and record observations.
Prior to beginning a test, the load cell and LVDT should be calibrated
to insure their linearity. In order to calibrate the load cell, it is
first removed from the test frame. A threaded rod is then attached to the
load cell and weights are suspended from the load cell. A voltage reading
is recorded at each load increment and at the conclusion of the calibration,
the points are plotted to insure a straight line between the points. The
LVDT calibration is conducted in a similar manner, except the rod of the
LVDT is pulled to various distances using a scale and the corresponding
voltage is recorded. Once again, the points are plotted to insure the
relationship is linear.
Once the power is turned on to the SSCTM, the load cell must be zeroed.
First, the specimen must be attached to the load cell. When the specimen
is attached to the load cell, the assembly is placed so the weight of the
specimen is hanging from the load cell. This procedure allows the load
45

cell to be zeroed with the weight of the specimen included. The load cell
is zeroed by adjusting the resistance of R4 on the signal conditioner board
and using the osc i 11 oscope to read the mean vo 1tage. The adjustab 1e
resistor (R4) on the signal conditioner board is shown in Figure 3.16.
The next step is to minimize the output signal to the servovalve.
Since the load cell is now reading zero load and the function generator
is not sending a signal, the output to the servovalve should be zero. This
output may be measured across pins 12 and 13 of the Servocontroller board
and is adjusted using the blue control of the function generator as shown
in Figure 3.17. If the output is not minimized, the load cell could be
damaged.

-
- a;na·•'ll •
'

Figure 3.16: Adjustable Resistor on the Signal Conditioner Board.


46

Figure 3.17: Calibration Control on the Function Generator.

After completing this step, the hydraulic power unit may be turned
on. The cylinder may move slightly, but should move slowly and only a
limited distance. The load cell and specimen assembly may now be mounted
into the SSCTM. In order to position the cylinder, the load cell should
be pressed to lower the cylinder or released to raise the cylinder. Once
the holes are aligned, the bolts may be inserted and tightened. This is
preceded by the placement of the spacers between the glass and steel clamp.
Once these are in place, the LVDT may be attached to the specimen.
The cylinder must be positioned at the point where the load cell reads
zero because some load may have been applied during the installation
procedure. This step is completed with the blue knob on the function
47

generator. The cylinder now moves when the knob is turned. After completing
this step, the initial voltages for the LVDT and load cell should be taken
and input to Labtech Notebook to begin the data acquisition program.
The final step necessary is to set the amplitude, mean, and frequency
to the desired levels. These values are set by adjusting the appropriate
knobs on the function generator. The mean is adjusted by using the blue
control shown earlier in Figure 3.17. The amp 1itude and frequency controls
are shown in Figure 3.18. The oscilloscope is used to ensure the proper
values are obtained as shown in Figure 3.19. Table 3.2 lists some common
stress levels and the corresponding peak voltage obtained from the
oscilloscope.
Once these steps are completed, the SSCTM should continue to run and
collect data until the specimen fails. The operator should make periodic
checks to insure the load function is being properly obtained. Minor
adjustments are occasionally necessary to keep the proper mean stress.
48

Figure 3.18: Amplitude and Frequency Controls on Function Generator.

Figure 3.19: The Oscilloscope Showing the Loading for a± 30 psi Test.
49

Table 3.2

Oscilloscope Readings for Selected Stress Levels

Stress Peak
Level Voltage
Reading
0 VPP
(psi) (Volts)

10 2.50
15 3.75
20 5.00
25 6.25
30 7.50
35 8.75
40 10.00
45 11.25
50 12.50
CHAPTER IV
TEST RESULTS

Introduction
This chapter presents the data that was obtained from the cyclic
testing of GE SSG 4000 structural silicone sealant specimens and the
analysis of the data. Each specimen was tested to failure which is achieved
when the specimen is no longer capable of sustaining the specified loading.
Also presented in this chapter is a comparison of these results to the
results of Sandberg and Rintala (1990).
The results include an S-N curve and the elongation time histories
for individual tests. The analysis consists of the Lilliefors Test for
Normality at each stress level and a regression analysis to fit an equation
to the data set. Finally, the analysis results and S-N curve are compared
to Sandberg and Rintala (1990) results.
Twenty-two specimens were placed under a full reversal sinusoidal
loading in order to construct the S-N curve. The number of cycles to
failure are shown in Table 4.1 and data for individual tests are shown in
Appendix C.

The Twentv psi Tests


The tests conducted on the structural silicone sealant using full
reversal sinusoidal loading with a 20 psi magnitude were the smallest
magnitude tests. A twenty psi stress corresponds to the design strength
of the GE Ultraglaze SSG-4000 structural silicone sealant.

50
51

Table 4.1
SSG Test Results

Test Test Cycles


Number Stress to
(psi) Failure
SSG-6 30 33500
SSG-7 40 28238
SSG-8 40 26100
SSG-9 30 37044
SSG-10 30 45486
SSG-11 40 30880
SSG-12 20 687600
SSG-13 20 392543
SSG-14 28 72448
SSG-15 25
SSG-16 40 11536
SSG-17 28 80339
SSG-18 40 49962
SSG-19 40 50120
SSG-20 40 33271
SSG-21 30 95901
SSG-22 30 13641
SSG-23 28 117810
SSG-24 28 92011
SSG-25 20 892067
SSG-26 20 216774
SSG-27 20 419832
52

The 20 psi tests are labeled SSG-12, SSG-13, SSG-25, SSG-26, and
SSG-27. The elongation time history for these tests with the exception
of SSG-25 are shown in Figures 4.1 thru 4.4, respectively. The elongation
time history from SSG-25 was lost due to a computer malfunction.
The time histories for the twenty psi tests are drastically different
than those at higher stress levels. The most noticeable difference is the
continually changing compressive elongation. While the average compressive
elongation remains virtually constant, the actual value fluctuates. The
initial tensile elongations were approximately 0.07 in. and slowly increased
during the first 20% of the specimen's life. Following this slow increase
is a stage where the tensile elongation remains almost constant for about
70% of the specimen's life. In the final stage, the tensile elongation
rises sharply as the specimen has substantial tears and the contact surface
area of the structural silicone sealant has been decreased significantly.
At this point, the specimen is no longer capable of sustaining the load
and the final tensile elongation is approximately 0.20 in. which is only
40% elongation.

The Twenty-eight psi Tests


The 28 psi tests are labeled SSG-14, SSG-17, SSG-23, and SSG-24. The
elongation time histories of the 28 psi tests are shown in Figures 4.5
thru 4.8. The stress level of 28 psi was selected because the first test
conducted at 25 psi had a large increase in the number of cycles to failure
SSG-12 20 psi
Elongation vs. cycles
0.1

0.05
,........_
...._.=
0
~
bO
·--=
0
=
fij
0

-0.05
0 50 100 150 200 250 300
Thousands
Cycles

Figure 4.1: Elongation Time History for SSG-12. (.TI


w
SSG-13 20 psi
Elongation vs. Cycles
0.2 . - - - - - - - - - - - - - - - - - - - - - - - - - - - - ,

0.15

0.1
""""
....~
'-"
~
0
~
0.05
bO
·--
~
0
fij
0

-0.05

-0.1
0 50 100 150 200 250 300 350 400
Thousands
Cycles

Figure 4.2: Elongation Time History for SSG -13. U'1


~
SSG- 26 20 psi
Elongation vs. Cycles
0.25

0.2

0.15

-..
'-"'
·-== 0.1
0
«<
·-.....
bO
=
0 0.05
m
0

-0.05

-0.1
0 50 100 150 200 250
Thousands
Cycles

(..11
Figure 4.3: Elongation Time History for SSG -26. (..11
SSG- 27 20 psi
Elongation vs. Cycles
0.2

0.15

0.1
-..
.e
""""'
c::
0
a:!
·-- 0.05
§
llJ
-
0

-0.05

-0.1
0 100 200 300 400 500
Thousands
Cycles

Figure 4.4: Elongation Time History for SSG-27. U1


0\
SSG-14 28 psi
Elongation vs. Cycles
0.4

0.3

0.2
·-c=_..r::
0
·-
-~
.Q 0.1
tiJ

-0.1
0 10 20 30 40 50 60 70 80
Thousands
Cycles

U'l
Figure 4.~: Elongation Time History for SSG - 14. .....,
SSG-17 28 psi
Elongation vs. Cycles
0.3 r - - - - - - - - - - - - - - - - - - - - - - - - - - - - ,

0.2

-.
~
.....
._..
~
0
·-.~.. 0.1
~
0
fil

0 ~-----------------------------------------------~

-0.1
0 10 20 30 40 50 60 70 80 90
Thousands
Cycles

U1
Figure 4.6: Elongation Time History for SSG-17. 00
SSG- 23 28 psi
Elongation vs. Cycles
0.4

0.3

-.
~ 0.2
~
·---
0
·-.~...
~
0
t- 1l 0.1

0 r---------------------------------------------------~

-0.1
0 20 40 60 80 100 120 140
Thousands
Cycles

(J"'
Figure 4.7: Elongation Time History for SSG -23 . 1.0
SSG- 24 28 psi
Elongation vs. Cycles
0.4

0.3

-.
c 0.2
'-"
·-
c
0
·-.~...
-r.LI§ 0.1

-0.1
0 10 20 30 40 50 60 70 80 90 100
Thousands
Cycles

0'1
Figure 4.8: Elongation Time History for SSG -24. 0
61

over the tests conducted at 30 psi tests. Therefore, the 28 psi stress
level was selected in order to attempt to better define the S-N curve in
this area.
The compressive elongations remain virtually constant throughout the
duration of the test at approximately -0.06 in. The initial tensile
elongations are approximately 0.08 in. and increase sharply during the
first 10,000 cycles. At this point, the rate of increase of the tensile
elongation slows until the final 20,000 cycles. The tensile elongation
rises abruptly at this point and continues until failure. The final tensile
elongation in each case was approximately 0.30 in. or 60% elongation.

The Thirty psi Tests


The 30 psi tests are numbered SSG-6, SSG-9, SSG-10, SSG-21, and SSG-22.
The final test, SSG-22, was excluded from statistical analysis because
numerous flaws found in the specimen were thought to result in an atypical
failure at 13,641 cycles. The 30 psi tests reveal a noticeable difference
from the 28 psi tests as the total number of cycles to failure decreases.
This decrease appears small, but considering only a 2 psi change in stress,
the difference is more significant. The structural silicone sealant is
designed for a stress of 20 psi and stresses of 30 and 40 psi are seldom
expected if ever.
The 30 psi elongation time histories are shown in Figures 4.9 thru
4.13. These time histories show the compressive elongations are virtually
constant throughout the entire duration of the test. The initial tensile
elongations were approximately 0.10 in. and increase significantly in the
SSG-6 30 psi
Elongation vs. Cycles
0.5

0.4 I-

0.3 I-
-.
~
_..
·-
~
0
0.2 I-
1:1:!
bO
·--
~
0
r.LI
-
0.1 I-

0
SSG-9 30 psi
Elongation vs. Cycles
0.4 . - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - ,

0.3 ~

-.
.5
_, 0.2 I-
s::
.9
....
as
~
0
fiJ 0.1 ~

0 ~------------------------------------------------~

-0.1
0 5 10 15 20 25 30 35 40
Thousands
Cycles

Figure 4.10: Elongation Time History for SSG -9. 0\


w
SSG-10 30 psi
Elongation vs. Cycles
0.4 . - - - - - - - - - - - - - - - - - - - - -- - - - - - - - - - - .

0.3

,........
0.2
·-_,c::c::
0
·-.~..
c::
0
UJ
- 0.1

0 r------------------------------------------------~

-0.1
0 10 20 30 40 50
Thousands
Cycles

Figure 4.11: Elongation Time History for SSG-10. 0'1


~
SSG-21 30 psi
Elongation vs. Cycles
0.3

0.25

0.2

0.15
r=
'-"
--·-
r=
0
CI:S
·-..... 0.1
bO
r=
0
r:ij 0.05

-0.05

-0.1
0 10 20 30 40 50 60 70 80 90 100 110
Thousands
Cycles

Figure 4.12: Elongation Time History for SSG -21. 0\


(.1'1
SSG-22 30 psi
Elongation vs. Cycles
0.25

0.2 r
0.15 ~

...-...
--------
'-"'
·-s::s:: 0.1 I-
0
~
·-...
bl)
s:: 0.05 I-
0
Ul
0

-o.o5 I ~

-0.1
0 2 4 6 8 10 12 14
Thousands
Cycles

Figure 4.13: Elongation Time History for SSG -22. 0'1


0'1
67

first 20,000 cycles of loading. After this increase, the tensile elongation
remains relatively constant until failure is imminent. As failure
approaches, the tensile elongation increases rapidly until the structural
silicone sealant is no longer capable of carrying the load. The final
tensile elongations vary, but typically range from 0.25 in. to 0.30 in.
which is 50% to 60% elongation.

The Forty psi Tests


The 40 psi tests are numbered SSG-7, SSG-8, SSG-11, SSG-16, SSG-18,
SSG-19, and SSG-20. One test, SSG-16, is excluded from statistical analysis
because numerous flaws found in the specimen were thought to result in an
atypical failure at 11,536 cycles. This particular specimen had an air
bubble at the contact surface between the structural silicone sealant and
the glass plate. This air bubble measured approximately 3/16 in. by 1/8
in. With the exception of one test, SSG-21, the total number of cycles
to failure of the 40 psi tests are virtually identical to that of the 30
psi tests.
The 40 psi elongation time histories are shown in Figures 4.14 thru
4.19. The elongation time history for SSG-20 is not shown. It was not
recorded due to a data storage malfunction during this test; however, the
test itself was not affected by this malfunction. With the exception of
SSG-16, the initial compressive elongations are approximately -0.06 in.
and remain virtually constant. The initial tensile elongations are
approximately 0.12 in. and slowly increase linearly until the final 25%
SSG-7 40 psi
Elongation vs. Cycles
0.25

0.2

0.15

-. 0.1
c
'-'
·-§
~
·-..... 0.05
bl)
c
0
~ 0

-0.05

-0.1

-0.15
0 5 10 15 20 25 30
Thousands
Cycles

Figure 4.14: Elongation Time History for SSG -7. C7'


(X)
SSG-8 40 psi
Elongation vs. Cycles
0.3

0.2 I-

-..
"-"
0.1 V
·-s::s::
0
·-~
-s::
0
fil 0

-o.1 r ------------------
__l _1_ _1_ __l
-0.2
0 5 10 15 20 25
Thousands
Cycles

Figure 4.15: Elongation Time History for SSG -8. 0'\


1.0
SSG-11 40 psi
Elongation vs. Cycles
0.3

0.2

,........
~
·-
~
0
0.1
~
-·--
~
0
~

-0.1
0
-----
5 10 15 20 25 30 35
Thousands
Cycles

Figure 4.16: Elongation Time History for SSG - 11. .....,


0
SSG-16 40 psi
Elongation vs. Cycles
0.4 , . . . . . . - - - - - - - - - - - - - - - - - - - - - - - - - - - ,

0.3 1-

0.2 1-
-.
.5
"'-"'
r::
0
·-.~.... 0.1 1-

r::
0
r.ll
-
0

-0.1 1-

-0.2
3 4 5 6 7 8 9 10 11 12
Thousands
Cycles

Figure 4.17: Elongation Time History for SSG-16. .....,


.....
SSG-18 40 psi
Elongation vs. Cycles
0.3

0.2

.........
.5
"-"

.§~ 0.1
§
til

0 r-------------------------------------------------~

-0.1
0 10 20 30 40 50
Thousands
Cycles

.......,
Figure 4.18: Elongation Time History for SSG-18. N
SSG-19 40 psi
Elongation vs. Cycles
0.4 . - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - . ,

0.3

.-.
.5 0.2 ·-
'-"
c
sa
~
~
0
fij 0.1

-0.1
0 10 20 30 40 50
Thousands
CycJes

Figure 4.19: Elongation Time History for SSG - 19. ......


w
74

of the specimen's life. During this final stage, the tensile elongation
increases rapidly until the structural silicone sealant is no longer capable
of carrying the load. Like the 30 psi test, the final tensile elongations
vary, but range from 0.25 in. to 0.30 in. which is 50% to 60% tensile
elongation.
Both the initial tensile and compressive elongations for SSG-16 are
approximately twice those of the other ± 40 psi tests. The compressive
elongation again remains virtually constant throughout the test duration.
The tensile elongation for this test increase 1inearly throughout the
entire test. However, after approximately 11,000 cycles, the tensile
elongation abruptly rises and failure occurred immediately thereafter.
The SSG-16 test also revealed another interesting characteristic that
is noticeable in other tests, but not as clearly. When the tensile
elongation changes, the compressive elongation also changes. This char-
acteristic in general holds true for all tests performed.

Defective Specimens
Three tests, SSG-16, SSG-22, and SSG-26, were reported in the test
data, but excluded from statistical analysis. Each of these tests failed
unusually quickly for their respective stress levels. Close inspection
of these specimens revealed the presence of several small voids or air
bubbles at the contact surface between the structural silicone sealant and
the glass and/or the aluminum block. These air bubbles reduce the surface
contact area of the structural silicone sealant and therefore increase the
stress level in the specimen during testing.
75

lilliefors Test for Normality


The Kolmogorov test (Conover, 1980), often referred to as the
Kolmogorov-Smirnov (KS) test, is based on statistical procedures which use
the maximum difference between a theoretical distribution function and an
experimental distribution function to test the null hypothesis that the
two distribution functions represent the same population. If the sample
size from which the experimental distribution function is taken is small,
the Kolmogorov test may be preferred over the chi-square test for goodness
of fit. Even for small samples, the Kolmogorov test is exact while the
chi-square distribution assumes the number of observations is large enough
that the sample provides a representative distribution of the data. For
most situations, the Kolmogorov test is more powerful than the chi-square
test, but there is a controversy over which test is more powerful in general
(Conover, 1980) .
In 1967, the first modification of the Kolmogorov test was made which
allows it to be used to test the hypothesis of normality. This modification
of the Kolmogorov test is called the lilliefors Test for Normality. The
null hypothesis states that the population is a normal distribution without
specifying the mean or variance of the distribution. When the lilliefors
test was first deve 1oped, a computer was used to acquire accurate estimates
of the exact distribution of the test statistic (Conover, 1980).
The procedure for the Lilliefors test has four general steps. The
only assumption required to use the Lilliefors test statistic is that the
76

data set comprises a random sample. Beginning with this random sample
data set, X1, Xz, .... , Xn, of an unknown distribution and denoted by F(x),

the first step is to compute the sample mean given by:

- (1) I:X
X=-
n
1\

t•l
1 • ( 4. 1)

This value is used as the estimate of the mean, ~ and is used to compute

the sample standard deviation as:

(4.2)

This value is used as the estimate of the population standard deviation,ar.

These two values are then used to compute the "normalized" sample values
Z1 which are used to compute the test statistic instead of the original
random samples. The "normalized" values Z1 are given by:

X-X
z 1 =-'-
5
(4.3)

The test statistic for the random sample is then computed by the
following:

T 1 =sup IF*(x)-S(x)l (4.4)


)(

where F*(x)denotes the specified hypothesized distribution function, and

S(x) denotes the empirical distribution function based on the random


77

sample. Equation (4.4) reads as 11


T equals the supremum, over all x, of
the absolute value of the difference F"(x)- S(x) 11 (Conover, 1980). The
null hypothesis, Ho, is rejected at the significance level, a., if the
appropriate test statistic, Tt, exceeds the 1-a quantile w1-o: as given by
Table 4.2.

Results of the Lilliefors Test for Normality


The null hypothesis, Ho, for each of the stress intervals is that the
experimenta 1 data is norma 11 y distributed, wh i 1e the a1tern at ive hypothesis,
H1 is that the experimenta 1 data is not norma 11 y distributed. These
hypothesis are simply stated as follows:
Ho: The data is normally distributed.
H1: The data is not normally distributed.

Based on the evidence, the null hypothesis, Ho, cannot be rejected at the
a= 0.05 significance level. Thus, the data at each stress level appears
to be normally distributed. The results of the Lilliefors Test are shown
in Table 4.3 and Figures 4.20 thru 4.23.
The fact that the data appears to be normally distributed is important
for future comparisons. The t-test and the Analysis of Variance (ANOVA)
are two common methods for comparing two samples. Both of these tests
assume the samples being compared are normally distributed. Thus, either
of these tests may be utilized to compare future test data to the results
of this testing.
78

Table 4.2

Quantities of the Lilliefors Test Statistic for Normality

n p

.80 .85 .90 .95 .99

4 .300 .319 .352 .381 .417


5 .285 .299 .315 .337 .405
6 .265 .277 .294 .319 .364
7 .247 .258 .276 .300 .348
8 .233 .244 .261 .285 .331
9 .223 .233 .249 .271 .311
10 .215 .224 .239 .258 .294
11 .206 .217 .230 .249 .284
12 .199 .212 .223 .242 .275
13 .190 .202 .214 .234 .268
14 .183 .194 .207 .227 .261
15 .177 .187 .201 .220 .257
16 .173 .182 .195 .213 .250
17 .169 .177 .189 .206 .245
18 .166 .173 .184 .200 .239
19 .163 .169 .179 .195 .235
20 .160 .166 .174 .190 .231
25 .140 .147 .158 .173 .200
30 .131 .136 .144 .161 .187
>30 .736 .768 .805 .886 1.031
.Jn .Jn .Jn rn rn
79

Table 4.3

Results of lilliefor Test for Normality

Test Mean Standard Standard Z(i) Cumulative


Stress Deviation Error of Probability
Mean X y

20 psi 598011 236967 118484 -0.867 0.250


-0.752 0.500
0.378 0.750
1.241 1.000

28 psi 90652 19809 9904 -0.919 0.250


-0.521 0.500
0.069 0.750
1.371 1.000

30 psi 52983 29050 14525 -0.671 0.250


-0.549 0.500
-0.258 0.750
1.477 1.000

40 psi 36429 10818 4416 -0.955 0.167


-0.757 0.333
-0.513 0.500
-0.292 0.667
1. 251 0.833
1.266 1.000
CUM FREQ Lilliefo~s Test fo~ No~Mality
1.9

I
.II 95~ LCL
/!'
.5
./'

95~ UCL I
9 2
-3·.9 -2".9 1."9 2."9 3".9

Figure 4.20: Lilliefors Test Results for the± 20 psi Tests. 00


0
CUM FREQ Lilliefo~s Test lo~ No~Mality
1.9

95Y. LCL

.5

95Y. UCL II
9 2
-3·.9 -2-.9 -1~9 9~9 1:9 2.-9 a·.9
.:.

Figure 4.21: Lilliefors Test Results for the± 28 psi Tests. co


.....
CUM FREQ Lilliefofs Test fof NofMality
1.9
.1"

-------'
NOR~(sJ
//
/
IJ 95Y. LCL
II
.5 /I' I // ---------
.... II ,/
--- II
------ ;"'
I/
95Y. UCL I
II
I
/'
/
........ I
________... I
/
9 z
-3-.9 -2-.9 -1~9 9:9 1.·9 2.·9 a·.9
(X)
Figure 4.22: Lilliefors Test Results for the± 30 psi Tests. N
C.UM FREQ Lilliefofs Test fof NofMality
1.9 T
!' I __.

I
//
t
I
t
!
.l I
:5Y. ~-
I I J"....
J""
--
/.1/ I .......
.5 + J
! .!"
I /
/~ I
./ /
I / ...
,•
/
I
/
!
LJ
7 /
..

// /
/
/ I
I
0 z
-3.0 -2.9 -1.9 9.9 1.9 2.9 3.9

Figure 4.23: Lilliefors Test Results for the± 40 psi Tests. 00


w
84

Previous Work
Previous work by Sandberg and Rintala (1990) resulted in the formulation
of an S-N curve for a Dow Corning neutral cure silicone sealant placed
under a full reversal sinusoidal load. The S-N curves, in semi-log and
hyperbolic form, for the previous work are shown in Figures 4.24 and 4.25.
Various methods are available to characterize and analyze the data,
but the most common form is a plot of failure stress versus the log of
cycles to failure. This equation is as follows:

a1 = a0 +A log N . (4.5)

where a 1 denotes the failure stress, N denotes the number of cycles to

failure, and A and o 0 denote constants to be estimated from regression


analysis. At zero stress, this equation implies a finite life span which
is only possible if the material is aging by a process which is not related
to load-induced stresses. If a finite life span can be determined, then
Equation (4.5) provides a reasonable estimation. A second method is to
characterize the failure stress versus the number of cycles to failure.
The equation used for this method is as follows:

(4.6)

where o 1 denotes the failure stress, N denotes the number of cycles to

failure, oa denotes the endurance limit predicted by regression, and K and


~ ~-----------------------------------------------------------,

-400

,......
a
~

- 300
~ I D [IJ
0::
""
t;
""a::
:> 200
.J

100

0 ~---------,----------~--------~----------r---------~--------~
0 2 4 &

LOG(CYCl£S TO FAA..URE)

Figure 4.24: S-N Curve in Semi-log Form for Sandberg and Rintala (1990) (X)
U'l
Data.
~ ~----------------------------------------------------------,

400

,....
a
Q.
.li(

-fa 300

w
It:
...
Ul
w
a:: 200
::>
...J
0
~
ID
D
100

0 ~--~----~--~----T----r----~--~--~----~--~----~---r--~
0 20 -40 60 8) 100 120
(Thou BOnds)
CYCL£S TO FAllJRE

Figure 4.25: S-N Curve in Hyperbolic Form for Sandberg and Rintala (1990) 00
Data. en
87

n denote material constants, which are either estimated by regression or


assumed. This equation is a hyperbolic fit when n = 1 which is commonly
known as the Wohler curve (Sandberg and Rintala, 1990). The resulting
equations and appropriate statistical data from Sandberg and Rintala are
shown in Table 4.4.
The structural silicone sealant used in this research, as well as that
tested by Sandberg and Rinta 1a ( 1990), was a neutra 1 cure structura 1
silicone sealant placed under a full reversal sinusoidal load. The general
shape of the S-N curve is almost identical to that of Sandberg and Rintala
(1990), but the number of cycles to failure at any given stress level is
much greater for these tests. A graphical comparison at various stress
levels between previous research (Sandberg and Rintala, 1990) and data
collected for this research is shown in Figure 4.26. This higher number
of cycles to failure may be attributed to technical advances in the silicone
industry over the past five to eight years.

Regression Analysis
Regression analysis was performed on the recorded data. This analysis
was performed to determine the equation of the semi-log and hyperbolic
curves through the data. Table 4.5 shows the pertinent data and Figures
4.27 and 4.28 show the resulting curves. These values were obtained using
a commercially available software program called Statistica, manufactured
by StatSoft, Inc.
88

Table 4.4

Results from Sandberg and Rintala

Regression Equation Obs. R2 Se Sc -


X
Type psi
(kPa)
Semi-log u,=63.95- 9.12 log N 14 0.90 10.8 6.1 4.54
(o 1 =441 - 62.9 log N) 14 0.90 1.57 0.88 4.54

Hyperbolic a 19.14 + 64090


1= t 14 0.82 2.10 8758 18500

(a = 132 + 442 ooo)


1 14 0.82 14.5 60400 18500
1

where: t =time, hrs.,


N = number of cycles,
R2 = goodness of fit parameter,
Obs. = number of observations used in regression,
Se = standard error of estimate,
Sc = standard error of coefficient,
x=mean oft or N values for observations.
SSG vs. S & R
700

600 -

500 -
Q)
I
....
;:I ~
:::: "0 I 400 r-
Cl:l s:::
~ Cl:l
~
....0~ ;:I
0
~ ~ 1 300 I-

a
I
200 I-

)C
100 I-
)/<
~
lXXX
AI
I~ ~
ICXXXXXX
0 ~
20 30 40
Failure Stress (psi)
~SSG Data ~ S&RData
*Sandberg & Rintala Data Estimateu

Figure 4.26: Graphical Comparison of SSG Data versus S and R Data. 00


\0
90

Table 4.5

Results from SSG Data

Regression Equation Obs. Rz Sa Sc -


X
Type psi
{kPa)
Semi -1 og a,= 97.5 - 13.54 log N 18 0.75 3.94 1.94 4.93

(a 1 =672.4- 93.4 log N) 18 0.75 27.2 13.4 4.93


4% 14%
700000
Hyperbolic a 1 = 19. 2S + 18 0.98 1. 73 77000 177000
N

( (J f = 133 + 483~000) 18 0.98 11.9 531000 177000


9% 11%

where: N = number of cycles,


R2 = goodness of fit parameter,
Obs. = number of observations used in regression,
Se = standard error of estimate,
sc = standard error of coefficient,
x=mean oft or N values for observations.
45 310

40 276

35 241

30 000 207
(f) (f)
(f) (f)
Q) Q)
._ 25 172 ._
~ ~

(f) ·- (f) 0
(f)
Q) 0... Q)o....
._ ~._..Y
::::J
-0 20 138 ·-::::J
0
LL LL

15 103

10 69

5 34

0 0
0 1 2 3 4 5 6

Log (Cycles to Failure)

\0
Figure 4.27: S-N Curve in Semi-Log Form for SSG Data. ......
45 310

40 276

35 241

CDO
30 207
(f) (f)
(f) (f)
Q) (])
"-- 25 172 "--
~ ~

(J) ·- (J) 0
Ul
(l) Cl a>CL
"-- L-..::t:
::::l 0 0
::::l
20 138
0 0
LL. LL.

15 103

10 69

5 34

0 0
0 200000 400000 600000 800000 1000000

Cycles to Failure

Figure 4.28: S-N Curve in Hyperbolic Form of SSG Data. ~


N
93

The results of the regression analysis reveal that the hyperbolic


function is much better than that of the semi-log function. The hyperbolic
function seems to predict the constant number of cycles at the high stress
levels as well as the wide range of cycles to failure at the design stress
of 20 psi. The hyperbolic equation predicts an endurance limit just below
20 psi.
CHAPTER V
CONCLUSIONS

Sulllllary
The general objective of this research was to investigate the effects
of cyclic loading on structural silicone sealant. This objective has been
fulfilled by the completion of several specific objectives. These specific
objectives include: (1) the design and construction of a cyclic testing
machine, (2) the collection of data to construct an S-N curve, (3) the
determination of equations defining the S-N curve, and (4) comparisons of
the S-N curves to previous research.

Conclusions
Structural silicone sealants have performed well since being introduced
into the construction industry. Previous testing as well as in-service
use has shown that structural silicone sealants are an excellent glazing
material when designed and installed properly. The results of this testing
reaffirm those previous conclusions.
Assuming an average of five major wind gusts per hour, as stated by
Sandberg and Rintala (1990), during a design life of twenty years, the
structural silicone sealant is subjected to 876,000 cycles. For this
design life, the hyperbolic equation predicts a failure stress of 20.0
psi, which is the equivalent of the design stress. For the same design
life, the semi-log equation predicts a failure stress of 17.0 psi, which
is 85% of the design stress.

94
95

The hyperbolic equation suggests that the current accepted design


strength of this product is sufficient. However, the semi-log equation
suggests that the current design strength may be too high. Consideration
must be given to the fact that the test specimens were subjected to a
continual cyclic loading and not subjected to periods of relaxation as is
experienced during its in-service 1 ife span. The structural silicone
sealant is capable of "healing" or regaining strength during relaxation
periods; thus, the structural silicone sealant would be expected to survive
more cycles under this condition. On the other hand, this testing does
not consider the effects of environmental conditions which may be detrimental
to the design strength of the product.
Although more data should be generated to increase confidence, the
results of this research are encouraging. When compared to previous work,
the results reported herein indicate an improvement in the strength under
cyclic loads of structural silicone sealants over that reported only a few
years ago. Of course, the best test for structural silicone sealants is
actual in-service use; these products have withstood that test.

Recommendations
If fatigue testing of structural silicone sealants is to continue,
some minor additions and improvements should be made to ensure the best
results from the SSCTM. These improvements will prolong the life of the
SSCTM, provide easier operation, and enhance the present capabilities. The
suggested modifications are as follows:
96

1. The purchase and installation of a Moog Accumulator which


accumulates and maintains pressure. This accumulation of
pressure dampens out extreme pressure changes in the system
as the load changes between extreme pressures.
2. The improvement of the existing clamping mechanism on the
test frame. This procedure will make operation of the SSCTM
much easier and allow moment and/or shear of various magnitudes
to be placed on the specimen.
3. The installation of a permanent LVDT attachment bar. Currently,
the LVDT must be connected at the initiation of each test
and occasionally the temporary attachment releases and aborts
the test.
4. The SSCTM shou 1d be re 1ocated to a more appropriate 1ocat ion.
Since the SSCTM is very sensitive to dust and particles, it
should be located in a laboratory area with less activity
and dust. A new location could also be necessary in order
to conduct some environmental testing which could be harmful
to persons without proper attire.
5. The SSCTM should be protected electrically with the use of
surge and spike protection or other appropriate circuitry.
This will reduce or eliminate the down time resulting from
electrical problems.
97

Future Testing
Future testing should investigate the effects of non-zero mean stress
loadings. These tests are needed not only to determine whether the zero
or non-zero loading is the worst case condition, but because structural
silicone sealants in some areas are exposed primarily to tensile forces.
The results of these tests could help conclude if the zero mean stress
level is indeed the worst case condition.
Future testing should also include the implementation of wind simulation
programs to investigate the behavior of structural silicone sealants under
actual wind loading conditions instead of simple sinusoidal loads. The
inclusion of pressure peaks and high frequency, low stress regions may
significantly effect the 1 ife span of the structural silicone sealant
specimen.
Finally, future research should include combined environmental-cyclic
loading conditions. Thus far, results such as this have not been investigated
and reported. However, environmental factors alone have been shown to
greatly reduce the strength of structural silicone sealants. The presence
of acid rain and ozone combined has drastically reduced the strength of
structural silicone sealants and could be an area of interest.
LIST OF REFERENCES

Bailey, J. B., Minor, J. E., and Tock, R. Wm., "Experiments Involving


Structurally Glazed Insulating Glass Units," Glass Research and Testing
Laboratory, Texas Tech University, Lubbock, TX., January 1989.
Boresi, A. P., and Sidebottom, 0. M., Advanced Mechanics of Materials, 4th
Edition, John Wiley &Sons, Inc., New York, 1985.
Conover, W. J., Practical Nonparametric Statistics, John Wiley & Sons,
Inc., New York, 1980
Klosowski, Jerome, Sealants in Construction, Marcel Dekker, Inc., New York,
1989.
Melander, T., "Saving Lakes," Informator A.B. Gothenburg, Sweden, 1989,
p. 101.

Morse, A. C., Electrohydraulic Servomechanisms, McGraw Hill, New York,


1963.
Panek, J. R., "Status of High-Rise Building Sealants in the U. S.," Building
Seal ants: Materials, Properties, and Performance, ASTM STP 1069, Thomas
F. O'Connor, editor, American Society for Testing and Materials,
Philadelphia, 1990.
Sandberg, L. B. and Albers, M. P., "Durability of Structural Adhesives for
Use in the Manufacture of Mobile Homes - Task VIII Sealant Test Report,
report for U.S. Dept. of HUD under contract no. H-2817, 1980.
Sandberg, L. B., Carbary, T. M., and Gil son, A. E., "Specimen Length Effects
on Mechanical Properties of a Silicone Elastomeric Sealant,• Science
and Technology of Glazing Systems. ASTM 1054, C. J. Parise, Ed., American
Society for Testing and Materials, Philadelphia, 1989, pp. 105-111.
Sandberg, L. B. and Rintala, A. E., "Resistance of Structural Silicones
to Creep Rupture and Fatigue," Building Sealants: Materials. Prop-
erties. and Performance. ASTM STP 1069, Thomas F. O'Connor, editor,
American Society for Testing and Materials, Philadelphia, 1990.
Schmidt, C. M., Schoenherr, W. J., Carbary, L. D., and Tak ish, M. S.,
"Performance Properties of Silicone Structural Adhesives,• Science and
Technology of Glazing Systems. ASTM 1054, C. J. Parise, Ed., American
Society for Testing and Materials, Philadelphia, 1989, pp. 22-45.

98
99

Swanson, John G., "Structural Glazing Continues to Grow in the Market


Place," Glass Digest, August, 1987, p. 108.
lock, R. Wm., "Temperature and Moisture Effects on the Engineering Properties
of Silicone Sealants," Building Sealants: Materials. Properties, and
Performance. ASTM STP 1069, Thomas F. O'Connor, editor, American Society
for Testing and Materials, Philadelphia, 1990.
lock, R. Wm., Chew, C. H., and Minor, J. E., "Determination of the Modulus
of Elasticity of Structura 1 Si 1icone Seal ants, • Report by Glass Research
and Testing laboratory, Texas Tech University, December 1987.
APPENDIX A
SPECIFICATIONS FOR
THE SSCTM PARTS

100
101

Specifications for LCF-50 Load Cell

Signal Output: 2 mV/V nominal


Linearity and Hysteresis: ± 0.15% Full Scale
Zero Balance: ± 2% maximum
Repeatability: ± 0.05% Full Scale

Compensated Temperature Range: 60°Ftol60°F

Temperature Effect of Zero and Span: ± 0. 005% Full Sea 1e I ° F

Bridge Resistance and Gage Type: 350 ohm foil


Excitation Voltage: 10.0 Vdc
Safe Overload: 150% Full Scale
102

Specifications for Moog 760-910 Servovalve

Supply Pressure: Minimum: 200 psi


Maximum: 3500 psi
Proof Pressure: 150% of supply pressure at P port
100% of supply pressure at R port
Recommended Viscosity of Fluid: 60 - 450 SUS @ 100° F
10 - 97 eSt @ 38 o c
Supply Filtration: 25 llffi
absolute or finer
1325 = 75
Operating Temperature: Minimum: -40 ° F
Maximum: 275 oF
Rated Flow Tolerance: ± 10%
Syn111etry: < 10%
Hysteresis < 3%
Threshold: < 0.5%
103

Characteristics of Flow Control Servovalves

Filtration: Servovalves require a very clean fluid for optimal operation.


The recommended cleanliness level is ISO DIS 4406 Code 14/11 to ensure
proper operation and a long, trouble-free life.
Rated Flow: Servovalves are rated at 1000 psi valve drop. The flow under
a no-load condition, QNL, will vary with supply pressure as shown in Figure
A.1. This relationship is given by:

where: QR = rated valve flow at 1000 psi drop


Ps = supply pressure
Flow-load Characteristics: Control flow to the load will change with
various combinations of load pressure drop and electrical input. These
characteristics closely follow the following relationship:

where: QNL =no-load flow


i = actual/rated current
Pv = (Ps - PR) - PL
= valve pressure drop
Ps = supply pressure
= return pressure
= load pressure drop
Internal leakage: There are two sources of internal leakage; first, flow
through the hydraulic amplifier (known as tare flow)_ which is relatively
constant and second, flow around the spool which varies with its position.
Maximum internal leakage occurs at null.
104

VALVE PRESSURE DROP (BAR)

,
200 - -
8 8N

v
L
/
500
V"
rJ~,... . ~ V"
100
,(;J
./: v ./:
~ _,;

. /~
v
e/ v
./
v _,;
60
~ () "~
y .... ./ / /

40
/
v .~.... '~
v
~
v / ./

~v
/ o.C
v. "/ / ,.,.
..... ~ "~....
30 ./: ./ ?'
~·~ .
""'" ./: I"'./: L
v ,.,.
~"~~ ~
~
a..
Q 20 .........
~
/
/
Vo ~./ a\.~
~'I
/"
~
/
/
s-
a..
~
70
~
~
0
....J 15 ~
/ . / I.-' . /
.....
"~.... """./ ~ / 0
....J
u.

""v ""'/
u. ./: ~'I./:

v
50 0
0
v w

~ ,.,.
w t-
t-

v
<(

"~....
<(
a: 10 ~ a:
v / 0
0
<(
0 / / ./: 30
<(
0....J
....J
0
z
v ~
v
~ ~\..~
" 0
z

5.0
/ .0.... ''/ L 20
v ~
c

/ ./:
/
15

~
v ~\..~ ~ "
3.0 ./
v -=-Q~ ~/

2.0
/ y ,.,.
10

/ ./ 7
~ ./
1.5 ./
V" ./ 5
/ 'b':J-...; V"
1.0 "~....
~
~
/
/ 3
0.7 /

0.5
v ~

8 8
2
VALVE PRESSURE DROP (PSI)
- N

Figure A.l: Change in Rated Flow with Pressure.


105

Spool Driving Forces: The maximum hydraulic force available to drive the
second-stage spool will depend upon the supply pressure, multiplied by the
end area of the spool.
Pressure Gain: A measure of the change in control port pressures as the
input current is varied about the zero flow point. Pressure gain is
measured against a blocked load under no load flow conditions. Normally,
the pressure gain exceeds 30% of the supply pressure for a 1% change in
the rated current and can be as high as 100%.
Null Bias: Input current to the servovalve required to adjust the output
to zero flow. Most Moog Controls valves have a mechanical adjustment which
allow the null bias to be externally adjusted.
Frequency Response: Servovalve frequency response will vary with signal
amplitude, supply pressure, and internal valve design parameters. The
typical response varies with supply pressure, as expressed by the change
in frequency of the 90° phase point, as shown in Figure A.2.
Step Response: Servovalve step response will vary with amplitude, supply
pressure, and internal valve design parameters.
SUPPLY PRESSURE (BAR)
50 100 150 200 250 300

1.5
I I I I I II
1.4 - Fp natural frequency at other pressures
- =

v
1.3 F natural frequency at 3000 psi (210 bar)
- ref
1.2
/
1.1 v
Fp
1.0 ~
F
ref 0.9 /
0.8
,...,v
0.7 /
0.6 /
0.5 /
0.4 I
I
0.3
0 1000 2000 3000 4000 5000
SUPPLY PRESSURE (PSI)

Figure A.2: Frequency Response Change with Pressure .


APPENDIX B
LABTECH NOTEBOOK
SETUP

106
107

Labtech Notebook Setup


for SSG Fatigue Testing

Number of Channels 9
Current Channel(s) [nor n.. m] 1
Channel Type Analog Input
Channel Name LVDT
Interface Device 0: Dash-8
Interface Device Channel Number [0 .. 143] 1
Input Range ± 5 v
Scale Factor 0.583
Offset Constant -1.708
Buffer Size 2048
Number of Iterations 1
Number of Stages [1 .. 4] 1
Sampling Rate, Hz 10.000
Stage Duration, sec. [0.0 .. 1.0E+08] 1. 73E+06
Start/Stop Method Normal
Trigger Channel 1
Trigger Pattern to AND [0 .. 255] 0
Trigger Pattern to XOR [0 .. 255] 0
Time Delay, sec. [0.0 .. 1.0E+08] 0.000
Analog Trigger Value 0.000
Analog Trigger Polarity High
Number of Samples to Save (Pretrigger) 0
108

Number of Channels 9
Current Channel(s) [nor n.. m] 2
Channel Type Analog Input
Channel Name Load Cell
Interface Device 0: Dash-8
Interface Device Channel Number [0 .. 143] 2
Input Range ±5v
Scale Factor -8.038
Offset Constant 4.20E-02
Buffer Size 2048
Number of Iterations 1
Number of Stages [1 .. 4] 1
Sampling Rate, Hz 10.000
Stage Duration, sec. [0.0 .. 1.0E+08] 1. 73E+06
Start/Stop Method Normal
Trigger Channel 1
Trigger Pattern to AND [0 .. 255] 0
Trigger Pattern to XOR [0 .. 255] 0
Time Delay, sec. [0.0 .. 1.0E+08] 0.000
Analog Trigger Value 0.000
Analog Trigger Polarity High
Number of Samples to Save (Pretrigger) 0
109

Number of Channels 9
Current Channel(s) [nor n .. m] 3
Channel Type Time
Channel Name Clock
Format sssss.sss
Buffer Size 2048
Number of Iterations 1
Number of Stages [1 .. 4] 1
Sampling Rate, Hz 10.000
Stage Duration, sec. [0.0 .. 1.0E+08] 1.73[+06
Start/Stop Method Normal
Trigger Channel 1
Trigger Pattern to AND [0 .. 255] 0
Trigger Pattern to XOR [0 .. 255] 0
Time Delay, sec. [0.0 .. 1.0E+08] 0.000
Analog Trigger Value 0.000
Analog Trigger Polarity High
Number of Samples to Save (Pretrigger) 0
110

Number of Channels 9
Current Channel(s) [nor n.. m] 4
Channel Type Analog Input
Channel Name LVDT
Interface Device 0: Dash-8
Interface Device Channel Number [0 .. 143] 1
Input Range ± 5 v
Scale Factor 0.583
Offset Constant -1.708
Buffer Size 2048
Number of Iterations 479
Number of Stages [1 .. 4] 3
Sampling Rate, Hz 10.000 5.56E-04 10.000
Stage Duration, sec. [0.0 .. 1.0E+08] 0 3600.000 5.000
Start/Stop Method Normal Normal Normal
Trigger Channel 1 1 1
Trigger Pattern to AND [0 .. 255] 0 1 1
Trigger Pattern to XOR [0 .. 255] 0 0 0
Time Delay, sec. [0.0 .. 1.0E+08] 0.000 0.000 0.000
Analog Trigger Value 0.000 0.000 0.000
Analog Trigger Polarity High High High
Number of Samples to Save (Pretrigger) 0 0 0
Ill

Number of Channels 9
Current Channel(s) [nor n.. m] 5
Channel Type Analog Input
Channel Name Load Cell
Interface Device 0: Dash-8
Interface Device Channel Number [0 .. 143] 2
Input Range ± 5 v
Scale Factor -8.038
Offset Constant 4.20E-02
Buffer Size 2048
Number of Iterations 479
Number of Stages [1 .. 4] 3
Sampling Rate, Hz 10.000 5.56E-04 10.000
Stage Duration, sec. [ 0. 0 .. 1. 0E+08] 0.000 3600.000 5.000
Start/Stop Method Normal Normal Normal
Trigger Channel 1 1 1
Trigger Pattern to AND [0 .. 255] 0 1 1
Trigger Pattern to XOR [0 .. 255] 0 0 0
Time Delay, sec. [O.O .. l.OE+08] 0.000 0.000 0.000
Analog Trigger Value 0.000 0.000 0.000
Analog Trigger Polarity High High High
Number of Samples to Save (Pretrigger) 0 0 0
112

Number of Channels 9
Current Channel(s) (nor n.. m] 6
Channel Type Time
Channel Name Clock
Format sssss.sss
Buffer Size 2048
Number of Iterations 479
Number of Stages [1 .. 4] 3
Sampling Rate, Hz 10.000 5.56E-04 10.000
Stage Duration, sec. [0.0 .. 1.0E+08] 0. 000 3600. 000 5.000
Start/Stop Method Normal Normal Normal
Trigger Channel 1 1 1
Trigger Pattern to AND [0 .. 255] 0 1 1
Trigger Pattern to XOR [0 .. 255] 0 0 0
Time Delay, sec. [0.0 .. 1.0E+08] 0.000 0.000 0.000
Analog Trigger Value 0.000 0.000 0.000
Analog Trigger Polarity High High High
Number of Samples to Save (Pretrigger) 0 0 0
113

Number of Channels 9
Current Channel(s) (nor n.. m] 7
Channel Type Analog Input
Channel Name LVDT
Interface Device 0: Dash-8
Interface Device Channel Number [0 .. 143] 1
Input Range ± 5 v
Scale Factor 0.583
Offset Constant -1.708
Buffer Size 2048
Number of Iterations 1
Number of Stages [1 .. 4] 3
Sampling Rate, Hz 10.000 10.000 IO.OOO
Stage Duration, sec. [O.O .. l.OE+08] 0 IO.OOO 5.000
Start/Stop Method Normal Trig On Normal
Trigger Channel I 1 I
Trigger Pattern to AND [0 .. 255] 0 0 I
Trigger Pattern to XOR [0 .. 255] 0 0 0
Time Delay, sec. [0.0 .. 1.0E+08] 0.000 0.000 0.000
Analog Trigger Value 0.000 1.000 0.000
Analog Trigger Polarity High High High
Number of Samples to Save (Pretrigger) 0 IOO 0
114

Number of Channels 9
Current Channel(s) [nor n .. m] 8
Channel Type Analog Input
Channel Name Load Cell
Interface Device 0: Dash-8
Interface Device Channel Number [0 •. 143] 2
Input Range ± 5 v
Scale Factor -8.038
Offset Constant 4.20[-02
Buffer Size 2048
Number of Iterations 1
Number of Stages [1 .. 4] 3
Sampling Rate, Hz 10.000 10.000 10.000
Stage Duration, sec. [0. 0 .. I. 0[+08] 0.000 10.000 5.000
Start/Stop Method Normal Trig On Normal
Trigger Channel 1 1 1
Trigger Pattern to AND [0 .. 255] 0 0 1
Trigger Pattern to XOR [0 .. 255] 0 0 0
Time Delay, sec. [0.0 .. 1.0E+08] 0.000 0.000 0.000
Analog Trigger Value 0.000 1.000 0.000
Analog Trigger Polarity High High High
Number of Samples to Save (Pretrigger) 0 100 0
115

Number of Channels 9
Current Channel(s) [nor n.. m] 9
Channel Type Time
Channel Name Clock
Format sssss.sss
Buffer Size 2048
Number of Iterations 1
Number of Stages [1 .. 4] 3
Sampling Rate, Hz 10.000 10.000 10.000
Stage Duration, sec. [O.O .. l.OE+08] 0.000 10.000 5.000
Start/Stop Method Normal Trig On Normal
Trigger Channel 1 1 1
Trigger Pattern to AND [0 .. 255] 0 0 1
Trigger Pattern to XOR [0 .. 255] 0 0 0
Time Delay, sec. [0.0 .. 1.0E+08] 0.000 0.000 0.000
Analog Trigger Value 0.000 1.000 0.000
Analog Trigger Polarity High High High
Number of Samples to Save (Pretrigger) 0 100 0
APPENDIX C
FATIGUE DATA FOR
SSG TESTING

116
117

Table C.1. SSG-6 Data


Load: ± 30 psi
Frequency: 1.05 Hz
May 14, 1992

Cycles Compressive Tensile Percent Tear Time


Elongations Elongations Tensile Length (hrs}
(in} (in} Elongation (in}
788 -0.04 0.10 20.0 0.22
2100 -0.04 0.10 24.0 0.58
4620 -0.04 0.12 24.0 1.28
6090 -0.04 0.12 28.0 1.69
8295 -0.04 0.14 28.0 2.30
11760 -0.06 0.22 44.0 3.27
13283 -0.06 0.22 44.0 0.14 3.69
18638 -0.06 0.23 46.0 5.18
21210 -0.06 0.24 48.0 0.35 5.89
24623 -0.06 0.24 48.0 6.84
28665 -0.08 0.25 50.0 7.96
30765 -0.08 0.26 52.0 0.97 8.55
31920 -0.08 0.27 54.0 8.87
32760 -0.08 0.29 58.0 9.10
118

Table C.2. SSG-7 Data


Load: ± 40 psi
Frequency: 1.00 Hz
May 19, 1992

Cycles Compressive Tensile Percent Tear Time


Elongations Elongations Tensile Length (hrs)
(in) (in) Elongation (in)
1829 -0.072 0.112 22.4 0.51
3660 -0.072 0.112 22.4 1.02
5500 -0.083 0.107 21.4 1.53
7317 -0.087 0.112 22.4 2.03
9150 -0.087 0.124 24.8 2.54
11000 -0.087 0.134 26.8 3.06
12825 -0.091 0.134 26.8 3.56
14630 -0.087 0.140 28.0 4.06
16475 -0.086 0.143 28.6 4.58
18300 -0.091 0.144 28.8 5.08
20125 -0.077 0.144 28.8 5.59
21950 -0.091 0.149 29.8 6.10
23800 -0.091 0.177 35.4 6.61
25625 -0.098 0.190 38.0 7.12
27450 -0.098 0.218 43.6 7.63
27460 -0.103 0.218 43.6 7.63
119

Table C.3. SSG-8 Data


load: ± 40 psi
Frequency: 1.00 Hz
May 20, 1992

Cycles Compressive Tensile Percent Tear Time


Elongations Elongations Tensile length (hrs)
(in) (in) Elongation (in)
13 -0.095 0.101 20.2 0.00
1830 -0.099 0.120 24 0.51
3664 -0.104 0.123 24.6 1.02
5487 -0.103 0.151 30.2 1.52
7318 -0.104 0.159 31.8 2.03
9145 -0.105 0.161 32.2 2.54
10976 -0.112 0.161 32.2 3.05
12812 -0.112 0.181 36.2 3.56
14633 -0.118 0.182 36.4 4.06
16461 -0.118 0.196 39.2 4.57
18298 -0.116 0.203 40.6 5.08
20120 -0.110 0.210 42.0 5.59
21975 -0.112 0.223 44.6 6.10
23775 -0.114 2.683 536.6 6.60
120

Table C.4. SSG-9 Data


Load: ± 30 psi
Frequency: 1.00 Hz
May 21, 1992

Cycles Compressive Tensile Percent Tear Time


Elongations Elongations Tensile Length (hrs)
(in) (in) Elongation (in)
3607 -0.071 0.125 25.0 1.00
7217 -0.068 0.120 24.0 2.00
10826 -0.067 0.122 24.4 3.01
14431 -0.071 0.128 25.6 4.01
18041 -0.072 0.141 28.2 0.15 5.01
21646 -0.070 0.150 30.0 6.01
25258 -0.068 0.157 31.4 7.02
28861 -0.073 0.179 35.8 8.02
32470 -0.070 0.196 39.2 1.00 9.02
36078 -0.073 0.287 57.4 10.02
36083 -0.071 0.280 56.0 10.02
121

Table C.5. SSG-10 Data


Load: ± 30 psi
Frequency: 1.00 Hz
May 22, 1992

Cycles Compressive Tensile Percent Tear Time


Elongations Elongations Tensile Length (hrs)
(in) (in) Elongation (in)
3610 -0.071 0.092 18.4 1.00
7217 -0.062 0.098 19.6 2.00
10825 -0.060 0.120 24.0 3.01
14435 -0.060 0.134 26.8 4.01
18039 -0.068 0.170 34.0 5.01
21651 -0.074 0.176 35.2 6.01
25253 -0.071 0.189 37.8 0.75 7.01
28861 -0.076 0.192 38.4 8.02
32468 -0.043 0.190 38.0 9.02
36075 -0.048 0.199 39.8 10.02
39685 -0.046 0.221 44.2 0.75 11.02
43294 -0.049 0.291 58.2 12.03
122

Table C.6. SSG-11 Data


Load: ± 40 psi
Frequency: 1.00 Hz
May 26, 1992

Cycles Compressive Tensile Percent Tear Time


Elongations Elongations Tensile Length (hrs)
(in) (in) Elongation (in)
3610 -0.071 0.189 37.8 1.00
7217 -0.062 0.175 35.0 2.00
10826 -0.060 0.200 40.0 3.01
14431 -0.060 0.200 40.0 4.01
18038 -0.068 0.219 43.8 5.01
21649 -0.074 0.226 45.2 6.01
25254 -0.071 0.253 50.6 7.02
28863 -0.076 0.275 55.0 8.02
123

Table C.7. SSG-12 Data


Load: ± 20 psi
Frequency: 1.00 Hz
June 9, 1992

Cycles Compressive Tensile Percent Tear Time


Elongations Elongations Tensile Length (hrs)
(in) (in) Elongation (in)
28791 -0.042 0.072 14.4 8.00
57570 -0.051 0.076 15.2 15.99
104745 -0.041 0.110 22.0 29.10
133525 -0.044 0.106 21.2 37.09
162310 -0.045 0.111 22.2 45.09
191092 -0.048 0.114 22.8 53.08
219871 -0.038 0.115 23.0 61.08
248655 -0.045 0.125 25.0 69.07
277438 -0.039 0.128 25.6 77.07
306220 -0.045 0.132 26.4 85.06
335001 -0.045 0.142 28.4 93.06
363781 -0.045 0.178 35.6 101.05
124

Table C.8. SSG-14 Data


load: ± 28 psi
Frequency: 1.00 Hz
June 15, 1992

Cycles Compressive Tensile Percent Tear Time


Elongations Elongations Tensile length (hrs)
(in) (in) Elongation (in)
3598 -0.048 0.083 16.6 1.00
7209 -0.066 0.104 20.8 2.00
10821 -0.054 0.127 25.4 3.01
14420 -0.037 0.133 26.6 4.01
18028 -0.050 0.140 28.0 5.01
21641 -0.037 0.145 29.0 6.01
25243 -0.055 0.147 29.4 7.01
28854 -0.055 0.150 30.0 8.02
32462 -0.048 0.152 30.4 9.02
36073 -0.047 0.158 31.6 10.02
39674 -0.048 0.158 31.6 11.02
43281 -0.057 0.162 32.4 12.02
46887 -0.045 0.166 33.2 13.02
50500 -0.048 0.173 34.6 14.03
54103 -0.045 0.182 36.4 15.03
57712 -0.037 0.193 38.6 16.03
61321 -0.039 0.200 40.0 17.03
64929 -0.041 0.218 43.6 18.04
68538 -0.040 0.239 47.8 19.04
72143 -0.039 0.299 59.8 20.04
125

Table C.9. SSG-16 Data


Load: ± 40 psi
Frequency: 1.00 Hz
June 22, 1992

Cycles Compressive Tensile Percent Tear Time


Elongations Elongations Tensile Length (hrs)
(in) (in) Elongation (in)
3598 -0.101 0.232 46.4 1.00
7208 -0.096 0.259 51.8 2.00
10813 -0.108 0.314 62.8 3.00
10815 -0.110 0.331 66.2 3.00
10818 -0.107 0.334 66.8 3.01
126

Table C.10. SSG-17 Data


Load: ± 28 psi
Frequency: 1.00 Hz
June 23, 1992

Cycles Compressive Tensile Percent Tear Time


Elongations Elongations Tensile Length (hrs)
(in) (in) Elongation (in)
1800 -0.069 0.088 17.6 0.50
3604 -0.069 0.091 18.2 1.00
7210 -0.074 0.108 21.6 2.00
9015 -0.074 0.110 22.0 2.50
10822 -0.070 0.110 22.0 3.01
12623 -0.063 0.114 22.8 3.51
14429 -0.067 0.112 22.4 4.01
18035 -0.073 0.112 22.4 5.01
19837 -0.073 0.117 23.4 5.51
21643 -0.070 0.114 22.8 6.01
23445 -0.074 0.117 23.4 6.51
27057 -0.070 0.118 23.6 7.52
30663 -0.076 0.125 25.0 8.52
32465 -0.066 0.120 24.0 9.02
34268 -0.074 0.127 25.4 9.52
37877 -0.073 0.120 24.0 10.52
39680 -0.072 0.127 25.4 11.02
43291 -0.076 0.125 25.0 12.03
45091 -0.074 0.131 26.2 12.53
46895 -0.074 0.134 26.8 13.03
48699 -0.084 0.149 29.8 13.53
50505 -0.072 0.152 30.4 14.03
52310 -0.072 0.167 33.4 14.53
55914 -0.076 0.172 34.4 15.53
57721 -0.073 0.179 35.8 16.03
59525 -0.072 0.182 36.4 16.53
61329 -0.07 0.179 35.8 17.04
64932 -0.076 0.189 37.8 18.04
66739 -0.064 0.201 40.2 18.54
68544 -0.064 0.209 41.8 19.04
127

Table C.10. (continued)

Cycles Compressive Tensile Percent Tear Time


Elongations Elongations Tensile length (hrs)
(in) (in) Elongation (in)
70344 -0.063 0. 211 42.2 19.54
72150 -0.073 0.221 44.2 20.04
73952 -0.073 0.228 45.6 20.54
75756 -0.077 0.236 47.2 21.04
77561 -0.072 0.236 47.2 21.54
79365 -0.074 0.269 53.8 22.05
128

Table C.11. SSG-18 Data


Load: ± 40 psi
Frequency: 1.00 Hz
June 24, 1992

Cycles Compressive Tensile Percent Tear Time


Elongations Elongations Tensile Length (hrs)
(in) (in) Elongation (in)
1802 -0.056 0.131 26.2 0.50
3603 -0.067 0.144 28.8 1.00
5407 -0.061 0.148 29.6 1.50
7212 -0.069 0.160 32.0 2.00
9018 -0.059 0.165 33.0 2.51
10818 -0.066 0.167 33.4 3.01
12623 -0.063 0.172 34.4 3.51
14430 -0.059 0.175 35.0 4.01
16230 -0.054 0.180 36.0 4.51
18035 -0.066 0.181 36.2 5.01
19840 -0.059 0.187 37.4 5.51
21644 -0.053 0.188 37.6 6.01
23447 -0.066 0.188 37.6 6.51
25251 -0.066 0.194 38.8 7.01
27053 -0.063 0.197 39.4 7.51
28860 -0.062 0.194 38.8 8.02
30660 -0.056 0.200 40.0 8.52
32464 -0.057 0.204 40.8 9.02
34269 -0.054 0.209 41.8 9.52
36074 -0.051 0.214 42.8 10.02
37877 -0.053 0.222 44.4 10.52
39679 -0.046 0.225 45.0 11.02
41486 -0.057 0.225 45.0 11.52
43287 -0.057 0.237 47.4 12.02
45090 -0.05 0.237 47.4 12.53
46894 -0.041 0.252 50.4 13.03
48698 -0.046 0.272 54.4 13.53
129

Table C.12. SSG-19 Data


Load: ± 40 psi
Frequency: 1.00 Hz
June 25, 1992

Cycles Compressive Tensile Percent Tear Time


Elongations Elongations Tensile Length (hrs)
(in) (in) Elongation (in)
1803 -0.055 0.106 21.2 0.50
3604 -0.065 0.122 24.4 1.00
5409 -0.060 0.126 25.2 1.50
7212 -0.055 0.129 25.8 2.00
9015 -0.059 0.125 25.0 2.50
12622 -0.056 0.139 27.8 3.51
14427 -0.063 0.143 28.6 4.01
16230 -0.050 0.135 27.0 4.51
18035 -0.048 0.146 29.2 5.01
19838 -0.053 0.152 30.4 5.51
25252 -0.043 0.169 33.8 7.01
27054 -0.053 0.173 34.6 7.52
28857 -0.040 0.175 35.0 8.02
30663 -0.036 0.182 36.4 8.52
32467 -0.043 0.183 36.6 9.02
34269 -0.036 0.191 38.2 9.52
37875 -0.036 0.195 39.0 10.52
39680 -0.028 0.218 43.6 11.02
41483 -0.025 0.225 45.0 11.52
43287 -0.015 0.232 46.4 12.02
45091 -0.012 0.248 49.6 12.53
46895 -0.014 0.297 59.4 13.03
130

Table C.13. SSG-21 Data


Load: ± 30 psi
Frequency: 1.00 Hz
June 29, 1992

Cycles Compressive Tensile Percent Tear Time


Elongations Elongations Tensile Length (hrs)
(in) (in) Elongation (in)
1800 -0.065 0.101 20.2 0.50
3603 -0.058 0.102 20.4 1.00
5410 -0.058 0.103 20.6 1.50
7212 -0.057 0.105 21.0 2.00
9016 -0.054 0.113 22.6 2.50
10818 -0.055 0.118 23.6 3.01
12623 -0.052 0.120 24.0 3.51
14426 -0.054 0.125 25.0 4.01
16231 -0.055 0.126 25.2 4.51
18037 -0.057 0.127 25.4 5.01
19839 -0.050 0.127 25.4 5.51
21642 -0.052 0.130 26.0 6.01
23446 -0.041 0.133 26.6 6.51
25252 -0.050 0.134 26.8 7.01
28858 -0.052 0.136 27.2 8.02
30660 -0.041 0.141 28.2 8.52
32465 -0.050 0.141 28.2 9.02
34271 -0.048 0.147 29.4 9.52
36075 -0.044 0.147 29.4 10.02
39681 -0.038 0.15 30.0 11.02
41484 -0.050 0.15 30.0 11.52
43287 -0.054 0.153 30.6 12.02
45090 -0.045 0.157 31.4 12.53
46895 -0.047 0.156 31.2 13.03
48701 -0.047 0.157 31.4 13.53
52307 -0.043 0.155 31.0 14.53
54110 -0.04 0.164 32.8 15.03
55915 -0.044 0.158 31.6 15.53
57718 -0.044 0.16 32.0 16.03
59523 -0.04 0.167 33.4 16.53
131

Table C.13. (continued)

Cycles Compressive Tensile Percent Tear Time


Elongations Elongations Tensile Length (hrs)
(in) (in) Elongation (in)
61326 -0.043 0.168 33.6 17.04
63130 -0.038 0.168 33.6 17.54
64935 -0.037 0.167 33.4 18.04
66736 -0.030 0.169 33.8 18.54
68540 -0.031 0.172 34.4 19.04
70344 -0.027 0.175 35.0 19.54
72148 -0.030 0.178 35.6 20.04
73951 -0.035 0.183 36.6 20.54
75755 -0.026 0.188 37.6 21.04
77562 -0.034 0.192 38.4 21.55
79363 -0.030 0.195 39.0 22.05
81169 -0.027 0.195 39.0 22.55
82970 -0.027 0.198 39.6 23.05
84777 -0.040 0.192 38.4 23.55
86578 -0.038 0.195 39.0 24.05
88382 -0.027 0.202 40.4 24.55
90187 -0.026 0.206 41.2 25.05
91990 -0.040 0.21 42.0 25.55
93793 -0.035 0.212 42.4 26.05
95597 -0.033 0.266 53.2 26.55
97400 0.874 174.8 27.06
132

Table C.14. SSG-22 Data


Load: ± 30 psi
Frequency: 1.00 Hz
July 1, 1992

Cycles Compressive Tensile Percent Tear Time


Elongations Elongations Tensile Length (hrs)
(in) (in) Elongation (in)
1801 -0.085 0.151 30.2 0.50
3605 -0.069 0.167 33.4 1.00
5408 -0.069 0.171 34.2 1.50
7211 -0.066 0.183 36.6 2.00
9015 -0.075 0.192 38.4 2.50
10819 -0.066 0.196 39.2 3.01
12623 -0.071 0.206 41.2 3.51
133

Table C.15. SSG-23 Data


Load: ± 28 psi
Frequency: 1.00 Hz
July 1, 1992

Cycles Compressive Tensile Percent Tear Time


E1 ongat ions Elongations Tensile Length (hrs)
(in) (in) Elongation (in)
1800 -0.062 0.090 18.0 0.50
3606 -0.059 0.097 19.4 1.00
5407 -0.055 0.100 20.0 1.50
7213 -0.060 0.105 21.0 2.00
9018 -0.058 0.105 21.0 2.51
10821 -0.056 0.108 21.6 3.01
12623 -0.060 0.108 21.6 3.51
14426 -0.065 0.111 22.2 4.01
16230 -0.058 0.111 22.2 4.51
18034 -0.065 0.114 22.8 5.01
19840 -0.060 0.115 23.0 5.51
21642 -0.062 0.114 22.8 6.01
23445 -0.069 0.114 22.8 6.51
25249 -0.060 0.117 23.4 7.01
27052 -0.066 0.118 23.6 7.51
28856 -0.066 0.118 23.6 8.02
30664 -0.060 0.118 23.6 8.52
32465 -0.053 0.118 23.6 9.02
34268 -0.063 0.118 23.6 9.52
36074 -0.060 0.118 23.6 10.02
37876 -0.060 0.122 24.4 10.52
39680 -0.059 0.122 24.4 11.02
41484 -0.060 0.121 24.2 11.52
43287 -0.055 0.131 26.2 12.02
48700 -0.065 0.135 27.0 13.53
50504 -0.059 0.135 27.0 14.03
52306 -0.063 0.136 27.2 14.53
57717 -0.058 0.139 27.8 16.03
61326 -0.053 0.145 29.0 17.04
64936 -0.062 0.150 30.0 18.04
134

Table C.15. (continued)

Cycles Compressive Tensile Percent Tear Time


Elongations Elongations Tens i1 e Length (hrs)
(in) (in) Elongation (in)
68540 -0.063 0.152 30.4 19.04
73952 -0.065 0.154 30.8 20.54
75755 -0.065 0.157 31.4 21.04
77559 -0.066 0.159 31.8 21.54
81167 -0.062 0.161 32.2 22.55
84777 -0.067 0.166 33.2 23.55
91991 -0.070 0.174 34.8 25.55
97400 -0.065 0.178 35.6 27.06
99205 -0.059 0.181 36.2 27.56
101010 -0.066 0.182 36.4 28.06
102817 -0.065 0.184 36.8 28.56
104619 -0.059 0.188 37.6 29.06
106420 -0.065 0.191 38.2 29.56
108225 -0.062 0.194 38.8 30.06
110028 -0.065 0.199 39.8 30.56
111833 -0.066 0.202 40.4 31.06
115439 -0.065 0.215 43.0 32.07
117244 -0.063 0.261 52.2 32.57
117245 -0.060 0.286 57.2 32.57
135

Table C.16. SSG-24 Data


Load: ± 28 psi
Frequency: 1.00 Hz
July 3, 1992

Cycles Compressive Tensile Percent Tear Time


Elongations Elongations Tensile Length (hrs)
(in) (in) Elongation (in)
1800 -0.060 0.089 17.8 0.50
3606 -0.056 0.095 19.0 1.00
5708 -0.057 0.099 19.8 1.59
7214 -0.058 0.104 20.8 2.00
9017 -0.058 0.105 21.0 2.50
10823 -0.059 0.108 21.6 3.01
12624 -0.060 0.109 21.8 3.51
14428 -0.060 0.113 22.6 4.01
16231 -0.059 0.114 22.8 4.51
18034 -0.061 0.116 23.2 5.01
19839 -0.062 0.116 23.2 5.51
21641 -0.064 0.118 23.6 6.01
23443 -0.066 0.119 23.8 6.51
25247 -0.061 0.119 23.8 7.01
27053 -0.066 0.120 24.0 7.51
28855 -0.065 0.120 24.0 8.02
30665 -0.062 0.120 24.0 8.52
32464 -0.057 0.121 24.2 9.02
34269 -0.060 0.122 24.4 9.52
36075 -0.059 0.122 24.4 10.02
37877 -0.058 0.123 24.6 10.52
39680 -0.055 0.125 25.0 11.02
41485 -0.054 0.126 25.2 11.52
43286 -0.055 0.132 26.4 12.02
48700 -0.057 0.136 27.2 13.53
50505 -0.057 0.137 27.4 14.03
52307 -0.058 0.139 27.8 14.53
57718 -0.057 0.141 28.2 16.03
61325 -0.058 0.146 29.2 17.03
64937 -0.061 0.152 30.4 18.04
136

Table C.16. (continued)

Cycles Compressive Tensile Percent Tear Time


Elongations Elongations Tensile length (hrs)
(in) (in) Elongation (in)
68541 -0.062 0.155 31.0 19.04
73953 -0.063 0.179 35.8 20.54
75755 -0.062 0.189 37.8 21.04
77558 -0.064 0.203 40.6 21.54
81166 -0.065 0.225 45.0 22.55
84776 -0.065 0.247 49.4 23.55
91990 -0.064 0.293 58.6 25.55
137

Table C.17. SSG-26 Data


Load: ± 20 psi
Frequency: 1.00 Hz
August 5, 1992

Cycles Compressive Tensile Percent Tear Time


Elongations Elongations Tensile Length (hrs)
(in) (in) Elongation (in)
3599 -0.070 0.044 8.8 1.00
7202 -0.053 0.098 19.6 2.00
10806 -0.064 0.093 18.6 3.00
14406 -0.055 0.091 18.2 4.00
18008 -0.068 0.097 19.4 5.00
21610 -0.045 0.098 19.6 6.00
25214 -0.070 0.098 19.6 7.00
28814 -0.072 0.101 20.2 8.00
32417 -0.050 0.104 20.8 9.00
39624 -0.054 0.102 20.4 11.01
43226 -0.058 0.104 20.8 12.01
46831 -0.054 0.103 20.6 13.01
50429 -0.074 0.104 20.8 14.01
54036 -0.064 0.107 21.4 15.01
57637 -0.061 0.108 21.6 16.01
61236 -0.072 0.110 22.0 17.01
64840 -0.067 0.110 22.0 18.01
68444 -0.058 0.111 22.2 19.01
72046 -0.053 0.112 22.4 20.01
75646 -0.061 0.112 22.4 21.01
82853 -0.072 0.114 22.8 23.01
90059 -0.070 0.114 22.8 25.02
93660 -0.075 0.115 23.0 26.02
100864 -0.070 0.116 23.2 28.02
104467 -0.077 0.117 23.4 29.02
108069 -0.078 0.117 23.4 30.02
115275 -0.068 0.117 23.4 32.02
118875 -0.058 0.118 23.6 33.02
122479 -0.070 0.118 23.6 34.02
126080 -0.055 0.119 23.8 35.02
138

Table C.17. (continued)

Cycles Compressive Tensile Percent Tear Time


Elongations Elongations Tensile Length (hrs)
(in) (in) Elongation (in)
129682 -0.062 0.119 23.8 36.02
133285 -0.060 0.120 24.0 37.02
136886 -0.064 0.121 24.2 38.02
140491 -0.071 0.121 24.2 39.03
144092 -0.053 0.122 24.4 40.03
147697 -0.053 0.123 24.6 41.03
151299 -0.051 0.124 24.8 42.03
158501 -0.052 0.125 25.0 44.03
162106 -0.060 0.125 25.0 45.03
169308 -0.050 0.126 25.2 47.03
172910 -0.057 0.131 26.2 48.03
176514 -0.060 0.132 26.4 49.03
180120 -0.067 0.133 26.6 50.03
183719 -0.074 0.134 26.8 51.03
187322 -0.064 0.135 27.0 52.03
194525 -0.070 0.161 32.2 54.03
198129 -0.053 0.171 34.2 55.04
201731 -0.067 0.174 34.8 56.04
205335 -0.084 0.175 35.0 57.04
212538 -0.08 0.185 37.0 59.04
216139 -0.061 0.221 44.2 60.04
216142 -0.058 0.228 45.6 60.04
139

Table C.18. SSG-27 Data


Load: ± 20 psi
Frequency: 1.00 Hz
August 8, 1992

Cycles Compressive Tensile Percent Tear Time


Elongations Elongations Tensile Length (hrs)
(in) (in) Elongation (in)
3599 -0.050 0.062 12.4 1.00
7202 -0.062 0.069 13.8 2.00
10804 -0.062 0.071 14.2 3.00
14407 -0.048 0.072 14.4 4.00
18009 -0.058 0.073 14.6 5.00
25214 -0.056 0.079 15.8 7.00
28817 -0.056 0.083 16.6 8.00
32419 -0.056 0.083 16.6 9.01
36022 -0.048 0.088 17.6 10.01
39623 -0.053 0.088 17.6 11.01
43225 -0.058 0.088 17.6 12.01
46829 -0.065 0.086 17.2 13.01
54034 -0.048 0.085 17.0 15.01
57634 -0.043 0.086 17.2 16.01
61237 -0.060 0.085 17.0 17.01
64841 -0.040 0.092 18.4 18.01
72047 -0.058 0.092 18.4 20.01
75650 -0.057 0.095 19.0 21.01
82853 -0.050 0.093 18.6 23.01
86456 -0.052 0.093 18.6 24.02
90056 -0.050 0.098 19.6 25.02
97262 -0.058 0.098 19.6 27.02
104466 -0.053 0.092 18.4 29.02
108067 -0.052 0.098 19.6 30.02
111672 -0.058 0.098 19.6 31.02
118875 -0.056 0.095 19.0 33.02
122478 -0.046 0.095 19.0 34.02
129684 -0.062 0.098 19.6 36.02
133287 -0.060 0.100 20.0 37.02
136890 -0.048 0.099 19.8 38.03
140

Table C.18. (continued)

Cycles Compressive Tensile Percent Tear Time


Elongations Elongations Tensile Length (hrs)
(in) (in) Elongation (in)
147694 -0.056 0.095 19.0 41.03
154902 -0.063 0.100 20.0 43.03
158503 -0.049 0.098 19.6 44.03
162107 -0.053 0.099 19.8 45.03
165706 -0.050 0.098 19.6 46.03
169312 -0.063 0.099 19.8 47.03
172910 -0.073 0.100 20.0 48.03
180115 -0.048 0.098 19.6 50.03
187321 -0.058 0.098 19.6 52.03
194526 -0.069 0.095 19.0 54.04
198130 -0.060 0.099 19.8 55.04
205333 -0.049 0.102 20.4 57.04
212540 -0.067 0.098 19.6 59.04
219745 -0.049 0.102 20.4 61.04
226947 -0.048 0.099 19.8 63.04
230549 -0.072 0.098 19.6 64.04
234151 -0.046 0.100 20.0 65.04
237756 -0.059 0.095 19.0 66.04
241356 -0.056 0.099 19.8 67.04
244961 -0.067 0.105 21.0 68.04
248561 -0.066 0.100 20.0 69.04
252167 -0.069 0.102 20.4 70.05
259368 -0.059 0.105 21.0 72.05
266575 -0.059 0.099 19.8 74.05
270176 -0.056 0.106 21.2 75.05
273777 -0.045 0.100 20.0 76.05
277382 -0.049 0.105 21.0 77.05
280985 -0.065 0.110 22.0 78.05
288190 -0.055 0.113 22.6 80.05
291791 -0.046 0.112 22.4 81.05
295394 -0.050 0.106 21.2 82.05
298995 -0.053 0.106 21.2 83.05
302600 -0.055 0.108 21.6 84.06
306201 -0.069 0.115 23.0 85.06
141

Table C.18. (continued)

Cycles Compressive Tensile Percent Tear Time


Elongations Elongations Tensile Length (hrs)
(in) (in) Elongation (in)
309805 -0.069 0.109 21.8 86.06
313404 -0.059 0.118 23.6 87.06
317006 -0.069 0.125 25.0 88.06
320609 -0.053 0.123 24.6 89.06
324211 -0.058 0.118 23.6 90.06
327813 -0.067 0.129 25.8 91.06
331416 -0.069 0.128 25.6 92.06
335022 -0.067 0.129 25.8 93.06
338622 -0.058 0.137 27.4 94.06
342224 -0.062 0.137 27.4 95.06
353033 -0.052 0.139 27.8 98.06
356634 -0.050 0.136 27.2 99.07
360235 -0.062 0.143 28.6 100.07
363838 -0.055 0.143 28.6 101.07
367441 -0.062 0.167 33.4 102.07
371045 -0.065 0.169 33.8 103.07
374646 -0.071 0.170 34.0 104.07
381849 -0.068 0.172 34.4 106.07
392650 -0.059 0.175 35.0 109.07
399854 -0.057 0.180 36.0 111.07
407059 -0.062 0.183 36.6 113.07

You might also like