ADEHESNE PROPERTIES OF
CAST POLYSULFIDE EMSTOMERS

T’hasix for 613 939m of M. S.
MSCMGAN STATE UNEVERSITY-
Cari H; Kcnkla
196?

0-169

This is to certify that the

thesis entitled

ADHESIVE PROPERTIES OF
CAST POLYSULFIDE EIASTOMEIRS

presented by

CARL H. KONKLE

has been accepted towards fulfillment
of the requirements for

Master of Science degree in Chemical Engineering

Q/W

Major professor

Date September 91, 1961

 

 

LIBRARY

Michigan State
University

 

ADHESIVE PROPERTIES OF

CAST POLYSULFIDE ELASTOMERS

By

Carl H. Konkle

AN ABSTRACT OF A THESIS

submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

MASTER OF SCIENCE

Department of Chemical Engineering

1961

ABSTRACT
ADHESIVE PROPERTIES OF
CAST POLYSULFIDE ELASTOMERS

by Carl H. Konkle

In recent years considerable interest has developed in the
adhesive properties of various materials. This interest has been
heightened by the spectacular rise in the use of adhesives in
industry.

Experimental work was carried out to investigate the adhesive
properties of cast polysulfide elastomer on various adherends.*
Specific interest develOped in the effect of geometric parameters
such as bond thickness and diameter. The effects of varying
adherends and surface roughnesses were also studied.

Distribution of stress across the adhesive-adherend interface
was also studied and a probable stress concentration diagram was
derived. .

Results of this work indicate that, for the polysulfide elastomer
bonded to the particular adherends used, the following conclusions may
be drawn: Bond strength increases as bond thickness and bond thick-
ness-to-diameter ratio decrease. Diameter plays an additional role
in that the prdbability of a flaw occurring at the periphery, where
fracture initiates, decreases as the diameter of the bond decreases
as predicted by the theory of flaws. The effect of increasing
surface roughness is to increase the bond strength up to a roughness

of approximately 50 pin.

 

* Adherend - term applied to any base material upon which an
adhesive is placed.

Carl Ho Konkle

Relative magnitudes of bond strength tested on steel, glass,
and oxidized steel were as would be expected from consideration of
the specific adhesion forces involved. An attempt was made to
determine the adhesive strength of the polysulfide elastomer-ammonium
nitrate bond. This attempt resulted in inconclusive data due to
the hygrosc0pic nature of the nitrate and its structural instability.
On the basis of this work the writer concludes that any attempt
to predict strength contributions by adhesion in elastomer - solid
matrices must consider the effect of the elastomer film.thickness

and the particle diameter within the matrix being considered.

ADHESIVE PROPERTIES OF

CAST POLYSULFIDE ELASTOMERS

By

Carl H. Konkle

A THESIS

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

MASTER OF SCIENCE

Department of Chemical Engineering

1961

‘4‘ "
‘ . 'f
f"

To Lori

whose patience and understanding during the writing of this
thesis was in truth an.examp1e of the saying "They also serve who

only stand and wait."

ACKNOWLEDGEMENT

The author wishes to express his thanks to Professor J. W.
Donnell whose interest, supervision, and suggestions made the
writing of this thesis possible.

Thanks are also due to Dr. Clement Tatro, Department of
Applied Mechanics, for his assistance in the area of mechanical
testing.

The author also wishes to express thanks to Mr. William B.
Clippinger for his assistance and advice in the construction of
the mechanical apparatus for experimentation.

The writer is further indebted to the United States Army

Ordnance Corps for the Opportunity to Obtain an advanced degree.

ii

Dedication . . . . . . .
Acknowledgement . . . .
Table of Contents . . .
List of Figures . . . .
List of Appendices . . .
Theories of Adhesion and
Survey of Previous WOrk
Experimental Procedures
Presentation of Results

Conclusions . . . . . .

TABLE OF CONTENTS

Statement

Suggestions for Further WOrk . . .

Bibliography 0 o o o o 0

Appendix-.......

iii

of Problem

Page

ii
iii

iv

16

35
1+6

51
53

Figure
Figure
Figure
Figure
Figure
Figure
Figure
Figure
Figure

Figure

Figure
Figure
Figure
Figure
Figure

Figure

$r o)

oo—qun

\O

11
12
13
14
15

16

LIST OF FIGURES

Sample Configuration "A? . . . . . . . .
Sample Configuration "B" . . . . . . . .

Block Diagram of Experimental Apparatus

Drive MOtOI' Power 0 o o o o o o o o o o
Transducer Power and Data Transmission .
Force Transducer . . . . . . . . . . . .

Gimbal Top Jaw and Sample Insert . . . .

Sample Alignment Block . . . . . . . . .

Typical Test Trace . . . . . . . . . . .

Polysulfide on Glass and Steel-

Configuration "A" . . . . . . . . .

Stress Distribution . . . . . . . . . .

Polysulfide on Steel-Configuration "B" .

Effect of T/D Ratio . . . . . . . . . .
Effect of Thickness . . . . . . . . . .
Surface Roughness Effects . . . . . . .

Polysulfide on Various Adherends . . . .

iv

Page
17
17

22
2h
31
33

36

37
38

#1
he
an

LIST OF APPENDICES

Page

Appendix A - Experimental Data'

Table

Table

Table

Table

Table

Table

Table

I ..

II -

III -

VII -

Adhesion of Polysulfide Elastomer

to Glass (Sample Configuration "A") . . 54
Adhesion of Polysulfide Elastomer

to Steel (Sample Configuration "A") . . 57
Adhesion of Polysulfide Elastomer

to Steel (Sample Configuration "3") . . 59
Adhesion of Polysulfide Elastomer

to Steel of Varying Surface Roughness
(Constant T/D Ratio). . . . . . . . . . 61
Adhesion of Polysulfide Elastomer

to Glass (Sample Configuration ”B” -

T/D = 0.125) . . . . . . . . . . . . . 62
Adhesion of Polysulfide Elastomer

to Ammonium Nitrate (Sample Configuration
"B" - T/D = 0.125) . . . . . . . . . . 62
Adhesion of Polysulfide Elastomer to
Oxidized Steel (Sample Configuration

"B" - T/D = 0.125) . . . . J . . . . . 62

Appendix B - Elastomer Technical Data . . . . . . . . . . 65

Force Conversion Graph . . . . . . . . . . . 66

THEORIES OF ADHESION AND STATEMENT OF PROBLEM

THEORIES OF ADHESION AND STATEMENT OF PROBLEM

One of the greatest paradoxes of modern technology is that of the
phenomenon of adhesion. Adhesion is a paradox in that we use it daily
in a.multitude of ways, from sealing letters to launching a space probe;
and yet we do not really understand the principles and laws which
govern it.

The use of adhesives can be traced back to the production of
laminated hunting bows by-Mongol nomads (19). The first serious study
of adhesion began during W0rld War I (6).

Any study of adhesion must necessarily begin with a definition
of adhesion and a brief discussion of the mechanisms by which it
functions. Adhesion is defined by Campbell (3) as "the attraction
existing between the molecules of substances at an interface." This
attraction can be divided into two general categories; first,
mechanical adhesion which is the purely'mechanical entanglement of
the adhesive molecules and the adherend surface molecules, and, secondly,
'specific adhesion. Specific adhesion is a.more complex phenomenon
involving the primary and secondary forces of attraction between
molecules. These forces have been investigated in some detail by
several investigators (l)(8)(22)(2h)(25). In summary of these investi-
gations the primary forces of attraction may be considered as those
involving polar bonds, covalent bonds, and metallic bonds. The secondary
forces involved are considered to be the Van der Waals' forces involving
the effects of orientation of permanent dipoles, the induction forces of
induced dipoles, and the dispersion forces due to shift of electron

positions. Staverman (2h) cites investigations which indicate that the

specific adhesion forces which control the total attraction at the
interface are those of the secondary forces, with the primary
valence type forces playing a very minor role.

Considering the forces mentioned above, it is possible to
analyze the mechanisms by which an adhesive joint can react to an
externally applied force. The first mechanism is that of "molecular
orientation" proposed by McBain (1%). This theory proposes that chains
of oriented molecules extend throughout the adhesive layer. This
orientation would be a maximum at the interface and decrease toward
the center of the adhesive layer where the effect of the adherend
material would be a minimum. If the theory of orientation is the
correct explanation for the strength of the bond then the adhesive
would break in the center of the layer. Bond strength should be inde-
pendent of diameter but dependent on the thickness of the layer. If,
instead, the adhesive breaks at the interface then the theory of
"molecular orientation" does not apply.

A second theory is that of differences in physical structure
proposed by Konstantinova and explained by Meissner (15). This theory
considers the possible variance of the crystal structure of the
adhesive as the layer thickness changes. It is obvious that this
theory could only apply to the types of adhesives which exhibit crystal
structure and this would thereby eliminate most high polymer adhesives.
If the theory is correct, bond strength would be a function of thick-
ness and an interface break would eliminate this theory as a controlling
factor.

Another possible reason for bond failure is the flaw theory

proposed by Bikerman (2) based on the statistical theory of flaws

developed by Griffith (ll)(2l). In essence, this theory states that
the statistical probability of the presence of flaws of sufficient
magnitude to cause initiation of failure decreases as the thickness
of adhesive decreases. It would seem.probab1e that the same
statistical reasoning would apply to a decreasing cross sectional
area. The flaw theory would explain bond failure occurring in the
adhesive layer and also at the interface.

A.further theory of bond failure has been proposed by Orowan and
Stager. This theory, as discussed by Epstein (9), considers a
reduction of plastic flow of the adhesive when it is confined between
two adherend layers. Bond failure would occur at a point of maximum
plastic flow. Strength would therefore be greater as the thickness
of the adhesive layer decreased causing restriction of plastic flow.
Interface failure of the adhesive bond would indicate that this theory
is incorrect as no plastic flow would occur at this point.

A.final theory of adhesive bond failure is based on the distribution
of internal stresses. This theory has been explored at some length by
Meissner and Bauldauf (15). Their investigations indicate that the
tensile strength of the adhesive bond is reduced by superimposed shear
stresses from three basic sources. The first of these sources is the
dimensional change or shrinkage of the adhesive on setting. This
dimensional change sets up shear stresses parallel to the face of the
adherend. The second source of shear stresses is present in the bond
if the thermal coefficient of expansion of the adhesive and adherend
differ greatly. This effect is particularly noticeable in the use of
thermoplastic adhesives. The third source of shear stresses is

developed during the application of force to the adhesive bond. This

shear is developed due to the difference in elastic modulus and Poisson's
ratio for the adhesive and adherend. The result of this is a lateral
contraction of the adhesive causing a "necking" of the bond. The
magnitude of the resultant shear is the vector component of the applied
force computed at the angle of departure from the direction of force
application. This angle is physically described by the adhesive as

it "necks" down. It is apparent from the above that the magnitude of
the shear stress parallel to the interface will be dependent on the

bond thickness and bond diameter. These two parameters establish the
angle of departure for a given adhesive material.

As the shear stresses are a.maximum at the interface it would
appear logical to assume bond failure would occur at or near the
interface.

Frocht (10) has considered a cross section of a butt joint adhesive
bond and states that the rectangular cross section exhibits geometric
similitude for equal ratios of thickness to diameter. He further states
that for bonds exhibiting this similitude equal external stress will
result in equal internal stresses. This would indicate that varying
the diameter of adhesive bonds having equal thickness-to-diameter
ratios should have no effect on the bond strength (17).

Eirich (7) has stated that bond strength may be dependent on the
diameter of the bond as well as the thickness-to-diameter ratio. He
has indicated that little information is available on the subject.

In summary, it can be stated that adhesive strength is dependent
on.mechanical adhesion, specific adhesion, and the geometry of the

particular bond involved.

Current interest in the field of adhesives has led to such
diversified applications as fishing rods, automobile bodies, boats,
architectural uses, and aircraft assemblya Ndlitary uses include
radomes, mine cases, depth charges, rocket launchers, missile packages,
aircraft primary structures, and solid pr0pellent rocket fuels. The
magnitude of the adhesive industry is indicated by the U. S. Tariff
Report for 1960 (25) which shows an approximate figure of #50 million
pounds of adhesive materials produced for that year. A.breakdown of
this figure in millions of pounds is as follows: Household adhesives,
10; Construction, 15; Furniture, 15; Packaging, 70; Plywood, 300; and
Rubberlike adhesives, “O. This trend toward the expanded use of
adhesive materials has led to an increased need for research in the
field of adhesive bonding.

STATEMENT OF THE PROBLEM The purpose of this work was to
investigate the adhesive properties of cast polysulfide elastomers.
An attempt was made to determine the contributions made to bond
strength by thickness, diameter, adherend, and adherend surface

condition.

SURVEY OF PREVIOUS WORK

SURVEY OF PREVIOUS WORK

The earliest systematic study of adhesion has been reported (6)
to have occurred in the aircraft industry during W0rld War I.

The first investigation of the effect of bond geometry was
conducted by Crow (A) in 1924. Experiments were conducted using one-
quarter inch diameter c0pper rods bonded with varying thicknesses of
eutectic solder. Nineteen samples having thickness ranging from 0.2
to 5.1 mm were tested. Results of these tests showed bond strength of
10.8 tons per square inch for the thinner samples and 4.0 tons per
square inch for the thicker samples. The trend of the intervening samples
“was for strength to increase as the thickness of the bond decreased.
No attempt at explanation of this phenomenon was made by the
investigator.

In 1925 McBain and Hopkins (11+) observed the effect of thickness
using a shellac-creosote adhesive between surfaces of mild steel.
Samples of three-quarter inch diameter were used. No attempt was made
to precisely control thickness. Results indicated that for very thin
films of adhesive a maximum adhesive strength in tension was obtained
of about 5000 pounds per square inch. Similar tests using thicker
bonds exhibited a strength of 3000 pounds per square inch. Further
tests were conducted to investigate the effect of surface roughness on
bond strength. The results of these experiments indicated that for
very rough surfaces the bond strength was much weaker than for similar
bonds prepared on smoother surfaces. The investigators attributed

this difference to an increase in average thickness due to the increase

in roughness.

In 1940, Bikerman (2) conducted experiments specifically designed
to explore the effect of bond thickness on tensile strength. An
adhesive layer of paraffin wax was used between a brass block and a
steel cylinder of 0.6 cm, diameter. The surfaces were prepared by
grinding with emery paper. The block and cylinder were then heated
and the wax melted between them. Loading of the bond was accomplished
by means of weights suspended from a chain which was looped through
a hole in the steel cylinder. The test results for a series of 100
bonds averaging 57 microns (2.25xlO-3in.) in thickness showed a mean
strength of 2h.6 Kg. per square centimeter (349 #/in.2). For a series

2in.) thickness a mean

of #0 bonds averaging 540 microns (2.13x10-
strength of 1h.85 Kg. per square centimeter (210.9 #7in.2) was obtained.
Bikerman attributed the thickness-strength effect to the statistical

flaw theory. He stated that the prObability of a flaw existing in the
adhesive layer decreased as the thickness of the layer diminished. .A
mathematical analysis included in this work indicated that the probability
rule could only account for approximately two-thirds of the effect noted.
The investigator suggested that the remainder of the effect could be

due to a difference in crystal structure of the adhesive layer caused

by differing cooling rates or rates of evaporation of solvents. He
indicated that thinner'bonds would show a smaller crystal structure

than the thicker bonds and would consequently be stronger. The
investigator's final conclusion was that the total explanation of the
thickness-strength phenomenon was due to the superposition of the flaw
theory and the crystal structure effect.

A very notable contribution to the understanding of bond strength

was made by Frocht (10) in 191+l. A photoelastic analysis was made of

lO

stresses in thin cross sections to determine the effect of thickness

on the state of stress. In the analysis of the results of these tests
the investigator concluded that for geometrically similar cross sections
equal loading would produce proportional stresses. Thus, bonds with
equal thickness-to-diameter ratios would react pr0portionally to

equal loading.

In l9u8 Meissner and Merrill (16) investigated a series of factors
influencing adhesive bond formation. Among the variables investigated
were thickness, temperature of bond formation, temperature of bond
testing, and thermal coefficient of expansion of the adhesive relative
to the adherend. Tests were conducted using steel cylinders of
approximately one inch diameter bonded with various thermoplastic
adhesives. Tensile tests were used at a loading rate of 6500 pounds
per square inch per second. Bonds were prepared.by insertion of a
layer of molten adhesive between the ends of the cylinders which had
been preheated to the temperature of the adhesive. The entire assembly
was kept aligned during cooling to the testing temperature.

To test the effect of relative coefficients of expansion the
investigators used an aSbestine filler to adjust the coefficient of
polystyrene to that of the steel adherend. Tests for the effect of
testing temperature were conducted using polymethyl methacrylate
adhesive bonded in a 0.1 inch layer between cylinders of emeryhground
steel.

The results of the experiments conducted indicated the following
effects. Strength increased as bond thickness decreased. Temperature
of formation had no effect on bond strength within the range investigated.

Strength increased as the testing temperature was decreased until the

11

second order transition temperature was reached. At this point the
strength decreased sharply and then proceeded upward through a
maximum and declined to zero. The effect of adjustment of the
thermal coefficient of expansion was to decrease the slope of the
thickness-strength curve.
The investigators concluded that the above effects were due to
a combination of four mechanisms. The first of these is that of
shear stresses developing across the face of the bond due to unequal
coefficients of thermal expansion for the adhesive and adherend. The
second explanation involves an inequality of elastic modulus and
Poisson's Ratio between adhesive and adherend. The third mechanism
proposed was that of the Theory of Flaws controlling the thickness-
strength variation. The final explanation offered was the orientation
effects Of the adhesive molecules as the interface is approached.
They stated that all four mechanisms were probably operating but that
the last two; the Flaw Theory and Orientation were most probably
controlling the effects noted during their investigation of the problem.
In 1950 Kraus and.Manson (l3) conducted research into the problem
of theoretical adhesion as Opposed to effective adhesion. They assumed
that the theoretical attractive potential between adhesive and adherend
was due to van der waals forces alone and computed a theoretical maximum
strength. They stated that this value was Of the same order of
magnitude as the cohesional forces in polymers as calculated.by Mark
and de Boer (13a, b). Experiments were conducted using varying'bond
thicknesses of polyethylene and polystyrene between steel drill rod
cylinders 0.75 inch in diameter. The surfaces were prepared by surface

grinding with a Norton 76 grit wheel. The ground surface was then coated

12

with a 1% solution Of the polymer and dried under vacuum. The polymer
bonds were formed by heating and.molding. The specimens were cooled
slowly to avoid thermal stresses. Testing was done at various
temperatures ranging from -24°C. to 82°C. The tensile tests were
performed using an Olsen testing:machines with a universal joint at
each end Of the specimen to assure proper alignment.

Results Of the above experiments were plotted as bond strength
versus thickness for various testing temperatures. Bond thickness
ranged from 0.025 mm to 0.3 mm, The curves exhibited the trend of
increasing strength with decreasing thickness as had been shown by
previous investigators. Extrapolation of these plots to zero thick-
ness resulted in a limiting value of adhesion which agreed very
favorably with the cohesional strength of the polymers tested.

The investigators concluded that any variable which would tend
to increase the shear component at the interface would promote a
rapid decline Of strength as the thickness of the bond increased.
They cited these variables as: coefficient of thermal expansion,
elastic modulus, and shear strength.

In 1951 Meissner and Baldauf (15) extended work done'by Meissner
and Merrill (16). Experiments were conducted using the following
adhesive-adherend combinations: Brass bonded with eutectic solder,
steel bonded with paraffin wax, and steel bonded with polystyrene.
The faces Of the steel cylinders were oxidized. Bonds were prepared
of various thicknesses by placing the adherend cylinders in a V’block
and heating to the melting. point Of the thermOplastic adhesive. The
adhesive was then placed between the ends of the cylinders and the

desired thickness established by movement of the cylinders. The

l3

adherend material was free to move as the adhesive conctracted upon
cooling. After cooling the excess adhesive was trimmed using a hot
knife. The resultant bonds were tested in tension using a standard
rate of force application. Bond thicknesses Obtained were between
0.002 and 0.300 inch. Sample diameters ranged from 0.46 to 1.125 inches.

To determine the effect of bond geometry on strength, plots Of
strength versus thickness were made using the various sample diameters.
The results Obtained indicated the known trend of increasing strength
with decreasing thickness. The lines representing various diameters
exhibited differing slopes. The investigators then correlated
strength versus thickness-diameter ratio as suggested by the photo-
elastic studies of Frocht (10). A reasonable correlation was Obtained
for various sample diameters within the narrow range tested.

It was noted during testing that failure Of the soldered joints
occurred at the alloy interface between the adhesive and adherend.

The wax-steel bonds broke near the interface leaving a thin film of
wax on the steel. Failure of the polyethylene-steel bonds occurred

in the adhesive layer for most samples. It was Observed, however, that
some of the bonds appeared to initiate failure at the interface on

the periphery and then the failure migrated to the adhesive layer
toward the center of the sample leaving an annulus of polyethylene

on the steel adherend.

In analysis of the results Obtained the investigators concluded
that the increase in strength with decreasing thickness was attributable
to the distribution of internal stresses within the adhesive bond.

They stated that for small values Of the thickness-to-diameter ratio

the amount of adhesive in the bond was small enough to preclude any

14

appreciable contraction and therefore the internal stress would be
essentially equal to the applied stress. For large values of the
thickness-diameter ratio the adhesive has reached a maximum tendency
to contract and the addition Of‘more thickness would not effect the
shear at the interface.

In 1955 Koehn (12) studied the strength behavior of Buna N-phenolic
and Thiokol (thermosetting) adhesives. Tensile test specimens were
prepared using round steel rods having a bond area of one square
inch. Various adhesive thicknesses were used. Samples were prepared
by bonding at a temperature of 400°F. under a pressure of 200 pounds
per square inch. Specimens under one mil thickness were bonded at
a pressure of 1000 pounds per square inch. Extreme care was taken to
insure proper alignment of the halves Of the test specimens.

Results of these tests indicated a confirmation of the thickness-
strength phenomenon. It was noted that for extremely thin films of
adhesive (below 0.003 inch) a sharp drop in strength occurred. Bond
failure varied from adhesional at thin films to cohesional at the
thicker films. This trend was noted for both the Buna Nephenolic and
{Duckol adhesives. It was also noted that failure in the Buna N-
phenolic bonds occurred initially at the periphery.

In conclusion the investigator stated that bond strength did not
Consistently vary inversely with'bond thickness. An Optimum film
thickness was indicated, and this thickness would depend on the

specific adhesive employed.

EXPERIMENTAL PROCEDURES

EXPERIMENTAL PROCEDURES

In order to determine the effect Of sample geometry, adherend
material and adherend surface condition tensile tests were conducted
using the following adhesive-adherend combinations: Polysulfide on
steel, glass, oxidized steel, and ammonium nitrate. The castable
polysulfide elastomer was chosen.because of its dimensional stability,
its thermosetting character, and the writer's personal interest in
its use in the solid pr0pellent field. The adherend.materials were
selected as they represent a large span of solid state Characteristics,
namely; an amorphous material, a metal, a.metallic oxide, and an
active oxidizer.

Test sample configurations were used as shown in Figures 1 and
2. Sample configuration "Afl‘was used initially as it was designed to
eliminate the problem of "necking" of the sample. It was thought that
this configuration would allow a controlled flow of material into the
adhesive joint and thereby reduce lateral contractions which tend to
produce shear stress across the interface. This configuration was
rejected due to inadequate support of the elastomer resulting in
peeling of the sample at the periphery. Sample configuration."B" was
then selected as it represents a conventional butt joint with the
attendent advantages of ease of preparation, ease of testing, and
greatest strength (5)(20). This configuration does have the dis-
advantage Of lateral contraction during loading.

The experimental apparatus used is shown in block diagram form
in Figure 3. Force required to rupture the adhesive bond was applied

by means of a special testing machine as shown in Photograph 1. This

16

l7

 

 

 

 

 

 

 

 

 

[—11 i

I e L

 

 

 

 

 

 

///////

FIGURE! -SAMPLE CONFIGURATION "A"

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

/ / / / 7 7
FIGURE 2 'SAMPLE CONFIGURATION "a"

 

 

18'

325.54 nfizmzimaxm no 23135 508-3ng

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Esd
«:3 x
concooom
amazement.
. raaam
_
_ uxd dd
_ _ocod
_
_ octmtcoi
mic—u SE 03.5 w .3300 motor—om
2.6 x r I I I I

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

l9

 

Photograph 1 - Testing Machine

2O

testing machine was designed to gain a maximum.sensitivity to applied
force not available using larger commercial testing machines.

In order to detect small differences in limited force ranges it
was necessary that this testing machine have a full scale range of
O-AO pounds force as Opposed to the average range of 0-50,000 pounds
found in commercial testing machines. Additional requirements were

as follows:
(1) close control of crosshead speed,
(2) short crosshead travel,

(3) ease of insertion of samples into mechanism.to
prevent prestressing,

(4) assurance of sample alignment during testing;

(5) simplicity of Operation to minimize uncontrolled
variables, and

(6) positive holding of samples during testing.

Force was transmitted from.the drive mechanism by means of a
vertical moving crosshead driven by a screw thread. This lead screw
was powered by a 2A volt D.C. electric motor connected as shown in
Figure 4. The crosshead speed was 0.2 inch per minute.

Measurement of the applied force was done by means of a force
transducer consisting of a balanced Wheatstone bridge of four SR-h
(FAP 12-12) l20-ohm strain gages connected electrically as shown in
Figure 5. The strain gages were arranged in such a fasion as to
record the total vertical loading placed on two end-supported
cantilever beams. The position of the strain gages in the Wheat-
stone bridge is such that only vertical force is recorded. Horizontal
forces, twisting moments, and temperature effects are cancelled out
by the bridge (18). The mechanical arrangement of the force transducer

is shown in Figure 6. The main body of the transducer was constructed

 

 

 

 

 

 

 

‘

field ‘\ ; JP”

 

 

 

 

 

fld. sw.
brushes A J
G vs" D’A
F 7“.
rev. 3w.

FIGURE 4-DRIVE MOTOR POWER

 

 

 

 

 

 

 

 

 

 

 

 

 

 

DC. PWR.[ HO VAC
RECORDER
ans-
ML
ducer Ylnput
Xlnpu‘t
00A,
\fiI'H
l.5v Hqucell

FIGURE 5-TRANSDUCER POWER 8 DATA TRANSN|SSION

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

W W7
Ell .. H l.
l. :l H 1? H : : J
Zia H 1 ’
L ‘L Li
Ill -’ ‘32: ’3
In
[—SR'4 STRAIN GAGES (4)
T__. 'l
3" lJ fl“ 3
3' 8 fi’ 8 D at
t
'0.
7;: "a;
t—v— 2% —-n

 

 

FIGURE 5 -FORCE TRANSDUCER

 

 

23

of hard aluminum while the cantilever beams were made of stainless
steel. The maximum total design force for this transducer was
calculated to be forty pounds.

Loading was applied to the sample rods by means of a vertical
”T” rod with knife edges placed between the ends Of the cantilever
beams of the force transducer. The free end of this rod is
terminated in a ball joint nut which swivels in the hanger Of the
gimballed top jaw of the testing machine. This gimbal and insert
collars for various sample diameters are shown in Figure 7.

The design of the gimbal top jaw allows freedom of motion in
all directions. This coupled with the ball joint in the gimbal
hanger insures pr0per alignment of the samples during application of
force. The gimbal was constructed of brass with steel drill rod
pivot pins. Sample inserts of brass were drilled and reamed to the
size of each sample rod diameter. Locking of the sample rod was
accomplished.by means of a steel ball bearing riding in a hole drilled
at four degrees to the sample rod hole and'breaking into it at its
midpoint. Downward.motion results in a lateral movement of the ball
toward the sample rod creating a wedging action which locks the sample
rod firmly in the sample insert.

Power supply, monitoring, and recording equipment are shown

in Photograph 2. The bottom deck of the panel houses the 28 volt
D.C. battery power source for the testing machine drive mechanism,
Directly in front of this battery pack is the power supply for the
transducer Wheatstone bridge. This power pack is an Opad Electric
Company Mbdel KM81B D.C. Power Supply. It has a variable voltage out-

put from.0-30 volts D.C. at 10 amperes. Supply to the transducer bridge

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

‘l
FT a
|_l
7., -\.- 2L
- a I'—
<———1"-—>‘
Igu "
:lt—‘l:—_’Jr U
uh 1 _ i u '3
n
l
2% t
3'“
‘ E

 

 

Drill 8 Ream To Sample Size

 

 

p---

 

 

 

 

 

FIGURET -GIMBAL TOP JAW a SAMPLE INSERT

 

 

25

 

Photograph 2 - Test Console

26

is 13 volts at 100 milliamps.

Recording of force data versus crosshead motion was done using a
Model 2A, MOsely Autograph XeY recorder. This apparatus is a two
axis, flat bed, graphic recorder. The manufacturer's stated accuracy
for this instrument is 1% using a.mercury cell battery as a
calibration voltage source. Input sensitivity for the Y axis ranges
from.5 millivolts to 100 volts. X axis sensitivity is 7.5 millivolts
to 150 volts. A physical record of data was Obtained on Keuffel and
Esser 358-5L, 10 x 10 to 1 inch graph paper. Force calibration was
accomplished by use of dead weights. Y axis travel was plotted
against pounds force. X axis travel is controlled by means of a 1.5
volt mercury cell battery voltage dropped across a 100 ohm potentiometer
which is geared to the vertical motion of the testing machine cross-
head. The resultant recorder trace is then an X axis plot of cross-
head motion versus the Y axis plot of force applied to the adhesive
bond.

The upper portion of the panel assembly contains the necessary
switches, monitoring meters, and indicator lights to control and
monitor test Operations. In addition, the extreme top portion of the
panel contains a test bridge and time base for operational checks of
the recorder circuits.

Test Operations were started with the preparation of the adherend
materials. Steel and glass rods were cut to a length of 2-1/4 inches.
The ends of these rods were then ground by means of chucking the rod
into a drill press and advancing the end face into a sheet of 2A0 grit
emery paper placed on the drill press bed. Polished surfaces for

effect of roughness tests were ground using a fine abrasive wheel mounted

27

on a standard surface grinder. Extremely rough surfaces were prepared
by lathe turning. Upon completion of surfacing the rods were carefully
washed with a Cleansing powder, rinsed several times, and dried using
compressed air. The rods were then degreased using toluene. After
final cleaning the surface roughness was determined using a Brush
Model BL~103 Surface Analyzer with a Model BLr106 R.M.S. meter as

a readout device. The Model PA2 pickup head was used for these tests.
Several random readings were taken Of each surface and an average
value obtained. The desired level Of roughness for all tests except
those designed to test the effect Of surface roughness was 35.5
microinches. Any rod not averaging the above figure was reprocessed
until a satisfactory reading was Obtained.

Ammonium nitrate rods were prepared by melting the nitrate
(M.P.-l70°c.) in a crucible on a hot plate and pouring the molten
nitrate into preheated soft glass vials. After solidification and
cooling the vials were then scored using a file and broken away from
the nitrate rod. The ends of the rods were then hand ground using
240 grit emery paper. No attempt was made to measure the surface
roughness Of the ammonium nitrate sample rods in view Of the fact
that the pickup needle Of the surface analyzer tended to score the
material giving a false reading. It was felt that the surface
roughness would be in the same range as the steel and glass rods
having been prepared in a like manner. All sample rods, including
glass and steel, were stored in desiccators upon completion of
preparation until used.

Casting Of the elastomers onto the steel and glass rods was

done by using a polyethylene casting block. Polyethylene was chosen

28

as the adhesive materials showed little or no tendency to stick.
Holes were drilled through the casting block the same diameter as
the rods. The final diameter of these holes was such that a snug fit
was Obtained on the rods. The end of each rod Opposite the prepared
surface was rounded slightly to allow the rods to be pressed into
the casting block without shearing away the polyethylene sidewalls.
Sample rods were inserted into the casting blocks from the top and
pressed through until the prepared surface was the desired distance
below the surface Of the casting block. This distance varied
depending on the desired sample thickness. Allowance was made for
the final cutoff to the exact thickness.

Casting of the elastomers onto the ammonium.nitrate sample rods
was accomplished by means of a cellophane tape dam.placed around
the prepared end Of the sample rod. The height Of this dam was
adjusted to obtain the desired thickness plus an allowance for final
cutoff.

Formulation and mixing of the elastomer was conducted as out-
lined in Appendix B of this work. The polysulfide elastomer used
was a polysulfide base rubber supplied by the Permalastic Products
Company as PX+45. The curing agent used was a lead base catalyst
supplied as PXéh6.

After'mixing, the liquid elastomer was poured into the molds
to the pr0per level. After casting, the samples were placed in a
curing oven. Temperature and length of cure were closely controlled.
All samples of the elastomer were cured in a like manner.

Samples were taken from the oven after the apprOpriate cure and

removed from the molds. Removal from the polyethylene molds was done

29

by pressing the samples out from the back side Of the mold. Clearing
of the nitrate samples was done by stripping away the cellophane
tape dam.

Proper thickness for all samples was Obtained by cutting the
elastomer layer to the desired dimension using a lathe. The bottom
sample rod with the adhesive cast to it was locked into the proper
size collet and the lathe micrometer stop zeroed with a thin, sharp
cutoff tool aligned at the adhesive-adherend interface. The tool
holder was then indexed away from the interface to the desired thick-
ness using the micrometer stop. The excess adhesive layer was then
cut away at high speed leaving a clean adhesive face at right angles
to the axis of the sample rod.

Upon completion of the above operation the bottom rod-adhesive
assembly was secured to the tOp rod using an epoxy resin adhesive.
The formation of this high strength bond insured that the sample
would break either in the elastomer layer or at the interface between
the elastomer and the bottom rod. The function of the tOp rod was
to provide a locking point for the gimbal top jaw. One end of these
rods was turned down to the diameter of the glass and nitrate samples.
The tOp rods used for the steel adherend samples were cut from the
same diameter steel drill rod. 3

Alignment of the entire assembly during the curing of the
epoxy was assured using an alignment block as shown in Figure 8.

The various diameter tOp rods were placed in the bottom.half of the
block, their ends coated with a very thin layer of epoxy and the top
half of the block containing the bottom.rod assemblies brought down

onto them. Alignment of the two halves of the block was maintained

 

 

 

 

 

30

/

—

 

Dowel Plns

 

Drlll a Ream Holes To Sample Slze

 

 

4......er
. rm
rundown.“
_
Hnnvuhwn
_
lllllernll
_ _ l-
HHHWHHWIL
ill-¢-l-
TllTlul
Ill.l..r|ll
_
Ill lll
lllrulll
ilhlr-d
llhrTll
. _
_ .
mix:
I]: I1:
ILL: I
{an

 

 

 

 

 

X...

 

4

FIGUREB‘SAMPLE ALIGNMENT BLOCK

 

 

31

by means of dowel pins located at Opposite corners. Any small excess
Of epoxy drained onto the sides of the top rods and therefore did not
affect the elastomer adhesive layer. After curing of the epoxy bond
the top half Of the alignment block was removed and the assembled
tensile test specimens removed from the bottom half for testing.

The initial step in testing was to secure the bottom jaw collar
on the adherend rod by means of a set screw. The entire assembly
was held in a verical position during this Operation and extreme care
taken to prevent stressing Of the test bond. The entire assembly was
then slid into the yoke Of the lower jaw Of the testing:machine while
the gimbal top jaw was held up so as not to touch the upper sample
rod. The gimbal containing the proper size insert was then lowered
over the top rod. The entire sample was then raised into the sample
insert until the bottom collar contacted the bottom jaw yoke. The
sample in this position was ready for the application Of force.

After ascertaining that all recording and drive Circuits were
activated the recorder was zeroed in both axes and force applied to
the test bond.by starting the crosshead drive motor. Data on cross-
head motion and applied force was transmitted electrically to the
recorder where a pen trace of the test was recorded. A.typical trace
is shown in Figure 9.

Upon completion of the test cycle the force data was taken from
the recorder sheets by measuring the peak excursion of each trace and
converting this dimension to force by means of the force conversion
graph shown in Appendix B. This force was then corrected for the
‘weight of the bottom rod and bottom jaw collar. By use Of the cross

sectional area for the particular sample the force per unit area

32

 

 

 

 

 

 

 

1 1

FIGURES 'TYPICAL TEST TRACE

 

 

33

required to rupture the adhesive bond was calculated. An average
value for a particular sample size and configuration was obtained

by taking the arithmetic mean of several test samples.

PRESENTATION OF RESULTS

PRESENTATION OF RESULTS

Initial experiments were carried out using polysulfide elastomer
bonded to glass rods of varying diameters using sample configuration
"A" (Figure 1). These experiments were repeated using steel drill
rod as the adherend. Results from these experiments are shown in
Figure 10. It was noted during the latter tests that the center of
the sample was cavitating on application of force due to a lack of
support for the polymer. This cavitation resulted in a bending moment
at the interface and subsequent peeling of the sample. It was concluded
after analysis of the configuration and data plots that the curves
shown represent a possible stress concentration plot for this partic-
ular configuration. Replots of these curves made by placing the
maximum stress at the periphery and minimum.stress at the sample
center resulted in a plot much like the results obtained by Mylonas
(17) in his photo-stress studies of cemented joints. A comparison of
these curves is shown in Figure 11. In view of the shortcomings of
sample configuration "AV it was decided to continue experimental
work using a standard butt joint specimen of configuration ”B"

(Figure 2). All subsequent tests were conducted using this config-
uration.

Tests on polysulfide adhesive and steel adherend were rerun using
the new sample configuration on six sample diameters. These tests
were run using thickness-tO-diameter ratios Of 0.125, 0.250 and 0.500.
Results are shown in Figure 12. It can be seen that adhesive strength
increases as the sample diameter decreases and as the T/D ratio

decreases. It should also be noted that the effect of decreasing

35

90+-

80—-

70“—

m
o
I

e¢.(esl)
o

40“

30—

STEEL

0 GLASS

 

I I l 1 1 I I
. .3 A .5 .6 .7
DIAMETER IIN.)

FIGURE IO -POLYSULFIDE ON GLASS AND STEEL' CONFIGURATION ”A"

STRESS CONCENTRATION FACTOR

37

 

 

 

DISTANCE FROM CENTER—>-

HCURE u “STRESS onsmlaunou

F/A (ESL)

quL

20"

ml-

 

38

 

L l I
.l .2 .3

FIGURE I2 -POLYSULFIDE ON

L l
.4 .5

DIAMETER (IN.l

.t

STEEL - CONFIGURATION "a"

39

diameter is more pronounced as the T/D ratio is decreased as shown
by the increasing slope Of the T/D lines. The preceeding effect is
shown more strikingly in the plot of T/D versus F/A for various
diameters as shown in Figure 13.

TO investigate the effect of thickness data was replotted as
F/A versus thickness Of the adhesive layer. This plot is shown as
Figure 1h. The effect of diameter is again apparent in this plot.
The effect Of thickness is in line with the results Obtained by
Meissner (l5)(l6) and others (2)(12)(13).

The next experimental series was conducted to determine the
effects Of surface roughness on adhesive strength. Tests were conducted
using 0.375 and 0.750 inch diameter steel adherends with polysulfide
adhesive. Samples were tested with 5.0, 35.5, and 60.0 microinch
surface roughness using a constant T/D ratio of 0.250. Results of
these experiments are shown in Figure 15. It can be seen that as roughness
increases the adhesive strength increases up to approximately 50 micro-
inches where a tendency to level off is noted. The effect Of sample
geometry is shown in the greater strength of the smaller diameter
sample. The results of these experiments appear to be in Opposition
to the effect predicted by Reinhart (22) who stated that an increase
in surface roughness should result in decreasing strength due to stress
concentrations at the surface discontinuities. It is true that for
adhesive materials which exhibit dimensional changes upon setting or
cooling that Reinhart's analysis is correct, however, the materials
used in this work were dimensionally stable on polymerization and there-
fore were not subject to the stress concentration effect. The increase

in strength is attributed to the increase in available Contact surface

F/A (RSI)

80—

60"—

50*—

20—-

 

  

O.I25

POLYSULFIDE ON STEEL

4

 

I I .lllrlllrllJllll

0.250 0.500
T/D RATIO

FIGURE I3 ‘EFFECT OF T/D RATIO

41

¢

0'

WP

mmmzxoih mo hummum- ¢_ manor.

“.2: mmwvzxoi...
N6 ..0 00.0 cod

No.0

 

2.32? _n_

 

ZBOIm mmmwmiSo con. Jmuhm zo moEISm>JOd

_

_

L

 

IO.

 

om

I'I'S'dl- v/a

FA 4931)

60 —
at
40-
30»-
20 r

IO-

 

#2

POLYSULFIDE 0N STEEL
T/D=0.250

20 30 4O 50 60

ROUGHNESS '(IJIN.)

FIGURE I5 “SURFACE ROUGHNESS EFFECTS

43

as the roughness increased. The leveling Off of the curves indicates
that this increase in contact area is not constant but tends to plateau
for some reason.

In order to investigate the effects Of various adherends on the
polysulfide adhesive, tests were conducted on glass, oxidized steel,
and ammonium nitrate adherends. A constant T/D ratio of 0.125 was
maintained. The results Of these tests are shown in Figure 16 compared
with the corresponding steel adherend plot from Figure 12. Attempts
to gain data on various diameters of ammonium nitrate were unsuccessful
due to cracking and instability of the smaller diameters. An additional
pr0blem.encountered was that of the hygrOSCOpic nature of the nitrate.
The figure Obtained from an average of three samples is shown as a
comparison but it is felt that it does not represent a true evaluation
of the adhesive strength due to the moisture picked up by the adherend.
The relative positions of the curves for glass and steel are as one
would predict based on the concept of specific adhesion. Steel,
representing a metal, would be expected to exert more secondary forces
in the adhesive bond than glass which is an amorphous material. The
relative positions of the plots for steel and oxidized steel are as
one would expect from.the work done by Taylor and Rutzler (25) on
theoretical adhesion of polyethylene to steel and oxidized steel.

Their investigation indicated a higher theoretical adhesion to the
metal than to the metal oxide.

The closeness of the glass and oxidized steel plots would indicate
that the Si-O-Si and Fe-O-Fe secondary force energies are approximately

equal in respect to their attraction to the elastomer tested.

F/A IP.S.I.I

T D = O.I25

 

70».-

60*-

U'l
O
I

 

30”

20»-

AMMONIUM NITRA:E\\

I I J I I I
.l .2 .3 .4 .5 .6

DIAMETER (IN)
FIGURE I6 -POLY$ULFIDE ON VARIOUS ADHERENDS

 

N—

CONCLUSIONS

CONCLUSIONS

In view of the experimental evidence Obtained in this work the

following conclusions with regard to polysulfide elastomer are made:

1.

Adhesive strength increases as the T/D ratio decreases for

a given diameter bond.

Adhesive strength increases as the bond thickness decreases.
Bond diameter has a significant effect on adhesive strength.
This effect is exhibited in two ways: first in the geometry
of the bond and second in that the bond strength increases for
decreasing diameters at fixed T/D ratios. It is felt that
there is only one logical explanation for the latter. As one
decreases the diameter of the bond the probability Of a flaw
occurring in the periphery decreases according to the Theory of
Flaws. Inasmuch as the maximum stress on loading is at the
periphery due to bond geometry and initation of fracture at

the periphery was noted through this investigation this appears
to be the only logical explanation for the diameter effect noted.
For the elastomer studied the adhesive strength increases as
the surface roughness increases up to approximately 50 pin.
roughness.

Stress distribution across the face Of the particular sample
configuration used appears to vary from sample center to
periphery with a concentration factor at the periphery of
approximately 5 times that at the center.

Specific adhesion of the polysulfide elastomer to the limited

variety of adherends tested appears to be in the relative

1+6

47

order one would expect from consideration of the secondary
bond forces involved.

Predicted strength contributions made'by adhesion in
elastomer - solid matrices would have to be based on
consideration Of the effects Of bond thickness and particle

diameter in the particular range of the matrix being considered.

SUGGESTIONS FOR FURTHER WORK

SUGGESTIONS FOR FURTHER WORK

It is recommended that the present work be extended to include
similar studies at various testing temperaturasto include a range from
-65°F. to 120°F. This temperature range should cover the normal
working conditions of cast elastomers.

It is further suggested that studies be conducted both theoretically
and experimentally to develOp a test specimen configuration which would
eliminate lateral contraction Of the adhesive layer during stressing.
It is felt that the configuration "A” tested in the initial phases of
this work could be improved to accomplish this purpose.

It is this writer's personal hOpe that this work will be extended
to include more of the elastomers and oxidizers used in the formulation
of solid pr0pellents. Adhesion between elastomer and oxidizer has been
recognized as a strong factor in the behavior of pr0pellent grains
subject to large deformations (23). PrOpellents loaded with a large
amount of oxidizer have very thin layers of elastomer material between
a large range of oxidizer particle sizes. It is felt that any attempt
to evaluate the effects of adhesion in the above matrix must include
consideration of the effects of elastomer layer thickness and particle
diameter. Accurate information on adhesion of elastomers and oxidizers
obtained in the manner of this work could possibly be extrapolated to
give a better picture of the contribution of adhesion to the overall

mechanical strength Of the pr0pellent matrix.

BIBLIOGRAPHY

10.

ll.

12.

BIBLIOGRAPHY

Alter, H., and Soller, W., "Molecular Structure as a Basis for

Adhesion," Ind. Eng. Chem., 50, 922 (1958).

 

Bikerman, J. J., "Strength and Thinness Of Adhesive Joints,"

.J. Soc. Chem, Ind., 60, 23 (19A1).

 

Campbell, W. G., "The Part Played by Fluids in Adhesion,"

Adhesion and Adhesives (Wiley, New York, 1954) p. 65.

 

Crow, T. B., "Some Properties of Soft Solder Joints,"

go SOC. Chemo Ind-‘9 B, 65 (1921+).

 

DeBruyne, N. A. and Houwink, R., Adhesion and Adhesives

 

(Elsevier, New York, 1951) p. 93.
Drew, R. B., "PhysiCO-Chemical Properties of Animal Glue,"

Adhesion and Adhesives (Wiley, New York, 195A) p. 193.

 

Eirich, F. R., Rheology Theory and Applications, V01. 3,

 

(Academic Press, New York, 1960) p. A92.
Eley, D. D., "Some Possible Surface Factors Involved in Joint

Strength," Adhesion and Adhesives (Wiley, New York, 1954),

 

p. 23.

Epstein, 0., Adhesive Bonding of Metals (Reinhold, New York, 195A)

 

p. 1H9.

 

Frocht, M; M., Photoelasticity vol. 1, (Wiley, New York, 1941)
p. 362.
Griffith, A. A., ”The Phenomenon of Rupture and Flow in Solids,"

Phil. Trans. Roy. Soc., 221A, 163 (1920).

 

Koehn, G. W., "Behavior of Adhesives in Strength Testing,"

Adhesion and Adhesives (Wiley, New York, 1954) p. 120.

 

51

 

-13.

13a.

13b.

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

51

Kraus, G. and Manson, J. E., "Adhesion of Polyethylene and

Polystyrene to Steel," J. Polymer Sci., 6, 625 (1951).

 

Mark, H., Cellulose and Cellulose Derivatives (Interscience,

 

New York, 1943) p. 1001.

deBoer, J. H., Trans. Faraday Soc.,_32, 10 (1936).

 

 

McBain, J. W. and Hopkins, D. G., "On Adhesives and Adhesive

Action," J. Phys. Chemo, 29, 188

 

(1925).

Meissner, H. P. and Baldauf, G. H., ”Strength Behavior of

Adhesive Bond," Trans. Am. SOC. Mech. Engrs., 13, 697 (1951).

 

Meissner, H. P. and Merrill, E. W., "Some Factors Affecting

Adhesive Bond Formation," A.§.T.Mt Bull., 151, 80 (1948).
Mylonas, C., "Experiments on Composite Models With Applications

to Cemented Joints," Proc. Soc. Exptl. Stress Anal., 12(2),

 

 

 

129 (1955).

Perry, 0. C. and Lissner, H. R., The Strain Gage Primer (McGraw-

 

 

Hill, New York, 1955) p. 209.

Perry, H. A., Adhesive Bonding of Reinforced Plastics (McGraw-Hill,

 

 

New York, 1959) p. 1.
mid.) p. 20.

Ibid., p. 54.

Reinhart, F. W., "Survey Of Adhesion and Types Of Bonds Involved,"

Adhesion and Adhesives (Wiley, New York, 1954) p. 9.

 

Smith, T. L., "ElastomeriC-Binder and Mechanical Property Requirements,’

Ind. Eng. Chema,‘52, 776 (1960).

II

Staverman, A. J., "Molecular Forces,

(Elsevier, New York, 1951) p. 18.

Adhesion and Adhesives

 

I

 

 

 

\ .
. e r. .
0‘ e
e 0 e
x
p» u‘ I a n‘ e I ‘7
I o I e o a
\ \
,. g o I -
\ \
p o O o . .
I I n e .'
a u - I I
\ \
. I I 0 e k ‘ .
.1 ' I
o . u .
o
A. ~ 0 o v -‘
- e .h‘ n‘
. n e
o , O .-
x
v v I.
I O
I ' ° 1
t I Q -~
. . e‘ a o e
. I: ' P‘ ' ‘ ' ’

52

25. Taylor, D. Jr. and Rutzler, J. E. Jr., "Adhesion Using

Molecular Models," Ind. Eng. Chemt, 59, 928 (1958).

 

26. U. 3. Tariff Commission, Synthetic Organic Chemicals, United

 

States Production and Sales, 1960. NO. 207 (U. S.

 

Government Printing Office, washington, D.C., 1961)

Table 16.

APPENDIX A. EXPERIMENTAL DATA

54

TABLE I

Adhesion of Polysulfide Elastomer to Glass
(Sample configuration "A")

 

 

 

Diameter (in.) Force (#) F/A (#/in.2) Average F/A (#/in.2)

0.098 0.33 41.8

0.098 0.26 32.5

0.098 0.30 38.4

0.098 0.29 36.8

0.098 0.45 58.5

00098 0028 35.1!"

0.098 0.51 66.7

0.098 0.22 27.5

0.098 0.59 78.0

0.098 0.30 38.4

0.098 0.31 39.1

0.098 0.27 34.2

0.098 0.33 41.8

0.098 0.24 29.2 47.1
0.155 0.84 43.2

0.155 0.86 44.2

00155 0059 29'9

0.155 0.63 31.8

0.155 0.99 50.8

0.155 0.97 49.5

0.155 0.76 38.8

0.155 0.73 37-0

0.155 0.81 41.2

0.155 0.50 24.5

0.155 0.75 38.4

0.155 0.50 24.5

0.155 0.92 47.2 '
0.155 0.99 50.7 39.5
0.190 0.63 21.3

0.190 0.71 23.9

0.190 0.85 29.0

0.190 0.80 27.4

0.190 0.98 33.5

00190 .1008 3700

0.190 1.04 35.6

0.190 1.28 44.2

0.190 0.84 28.5

0.190 0.65 22.1

0.190 1.08 37.0

0.190 1.08 37.0

0.190 1.35 46.8

0.190 1.38 47-7

0.190 1.12 38.5

0.190 1.23 42.5 34.5

55

TABLE I - (continued)

Diameter (in:) Force (#) F/A (#/in.2)

Average F/A (#/in. 2)

0.221 1.38 35.4
0.221 0.72 18.1
0.221 0.99 25.1
0.221 1.20 30.6
0.221 0.90 22.6
0.221 0.89 22.4
0.221 0.54 13.4
0.221 ' 0.90 22.6
0.221 1.32 33.8
0.221 0.91 23.2
0.221 0.91 23.2
0.221 1.40 35.7
0.221 1.40 35.7
0.221 1.65 42.3
0.221 0.96 24.3 28.1
0.243 1.14 23.8
0.243 1.15 25.5
0.243 0.80 16.5
0.243 1.36 28.5
0.243 0.86 17.6
0.243 0.90 18.5
0.243 1.16 25.6
0.243 1.36 28.5
0.243 0.77 15.7
0.243 0.77 15.7
0.243 1.36 28.5
0.243 0.89 18.4
0.243 1.11 23.1
0.243 0.90 18.5
0.243 1.31 27.6 22.7
0.375 1.42 11.9
0.375 1.41 11.8
0.375 1.06 8.7
0.375 1.24 10.3
0.375 1.69 14.4
0.375 1.45 12.2
0.375 1.92 16.4
0.375 1.41 11.8
0.375 2.18 18.7
00375 1070 lit-)4-
0.375 1.41 11.8
0.375 1.50 12.6
0.375 1.45 12.2
0.375 0.74 6.3
0.375 1.76 14.9
0.375 1.58 13.3 13.6

56

TABLE I - (continued)

Diameter4(in.) Force (#) F/A (#/in.2) Average F/A (#[in.2)

0.504 2.52 12.2
0.504 2.97 14.4
0.504 1.42 6.7
0.504 2.23 10.7
0.504 3.34 16.2
0.504 1.98 9.4
0.504 3.13 15.2
0.504 2.29 11.1
0.504 2.10 10.1
0.504 2.33 11.2
0.504 2.67 12.9
0.504 3.28 15.9
0.504 1.74 8.2
0.504 3.34 16.2
0.504 2.34 11.2
0.504 2.72 13.2 12.7
0.735 3.44 7-9
0.735 4.02 9.2
0.735 3.42 7.8
0.735 3.25 7.4
0.735 3.08 7.0
0.735 3.74 8.6
0.735 4.79 11.1
0.735 3.84 8.8
0'735 30u6 7'9
0.735 3.30 7-5
0-735 2.90 6.6
0-735 3-90 8-9
0.735 4.42 10.2
0.735 4.77 10.9
0.735 3.70 8.5 8.8

57

TABLE II

Adhesion of Polysulfide Elastomer to Steel
(Sample Configuration "A")

 

 

 

Diameter fin.) Force (79‘) F/A (#/in.2) Average F/A (#jin.2)
0.0938 0.48 65.2
0.0938 0.40 53.8
0.0938 0.47 64.7
0.0938 0.54 74.4
0.0938 0.55 76.5
0.0938 0.45 62.2
0.0938 0.37 50.3
0.0938 0.44 60.4
0.0938 0.51 70.0
0.0938 0.53 73.5
0.0938 0.50 69.2
0.0938 0.44 59.1
0.0938 0.41 57.0
0.0938 0.44 59.1
0.0938 0.52 71.0 69.2
0.1562 0.93 46.7
0.1562 0.69 34.0
0.1562 0.94 47.0
0.1562 0.67 33.0
0.1562 0.93 46.7
0.1562 1.09 55.0
0.1562 1.19 60.0
0.1562 0.73 36.2
0.1562 0.74 36.4
0.1562 1.08 54.5
0.1562 1.37 69.7
0.1562 1.41 72.0
0.1562 0.91 45.3
0.1562 0.85 42.2
0.1562 0.66 32.5
0.1562 1.37 69.4 50.9
0.1875 1.39 48.7
0.1875 1.20 41.9
0.1875 1.55 54.5
0.1875 0.74 25.5
0.1875 0.98 34.0
0.1875 0.98 34.0
0.1875 1.12 39-3
0.1875 1.02 35.6
0.1875 1.60 56.5
0.1875 1.81 64.0
0.1875 1.81 64.0
0.1875 1.29 45.3
0.1875 1.04 36.3

58
TABLE II - (continued)

Diameter (in.) Force (#) F/A (#/in.2) Average F/A.(#/in.2)

 

0.1875 1.71 60.5
0.1875 1.22 42.9
0.1875 1.59 56.2 47.7
0.250 1.65 32.6
0.250 1.54 30.3
0.250 1.33 26.3
0.250 1.90 37.7
0.250 1.19 23.3
0.250 1.73 34.0
0.250 1.98 39.4
0.250 1.54 30.4
0.250 2.13 42.4
0.250 1.80 35.7
0.250 1.74 34.4
0.250 1.47 28.9
0.250 1.40 27.7
0.250 1.53 30.2
0.250 1.63 32.2
0.250 2.11 42.0 33.9
0.375 2.44 21.1
0.375 2.48 21.4
0.375 1.59 13.4
0.375 3.22 28.1 22.3
00500 3'75 1900
0.500 2.81 14.1
0.500 4.70 23.9
0.500 3.21 16.4 19.1
0.750 8.70 19.2
0.750 4.04 8.7
0.750 6.23 13.7
0.750 7.03 15.5 14.2

Diameter (in.)

59

TABLE III

Adhesion of Polysulfide Elastomer to Steel

(Sample Configuration "3")

 

 

0.1562
0.1562
0.1562

0.1875
0.1875
0.1875

0.250
0.250
0.250
0.250

0-375
0-375
0-375
0.375

0.500
0.500
0.500
0.500

0.750
0.750
0.750
0-750

Diameter (in.)
0.1562

0.1562
0.1562

 

0.1875
0.1875
0.1875

0.250
0.250
0.250
0.250

 

 

Average F/A (#/in.2)

 

 

A. T/D = 0.125

Force (#) F/A(#/in.2)
1.70 88.6
1.56 81.3
1.61 83.8
2.29 82.7
2.12 76.6
2.18 78.8
3.44 70.0
3.54 72.0
3.82 77.8
4.39 89.5
7.87 70.9
7.14 64.3
7.62 68.6
6.87 61.8
11.88 60.6
9.03 46.1
10.52 53.7
11.58 59.0
20.12 45.8
13.34 30.3
12.77 28.9
13.32 30.2
B0 Ell/D = 00250

Force (#) F/A.(#/in.2)
1.04 54.3
0.98 51.0
0.88 45.8
1.12 40.5
1.40 50.5
0.98 35.7
1.08 39.0
2.32 47.3
2.50 50.9
2.79 56.9
2.34 47.7

84.6

79-3

77-3

66.4

54.8

33-8

Average F/A (#/in.2)

 

50.4

41.4

50.6

60

TABLE III - (continued)

Diameter (in.) Force (#2 F/A (#/in.2)

Average F/A (#/in.2)

 

 

0.375 3.82 34.4

0.375 4.02 36.2

0.375 3.29 29.6

0.375 4.51 40.6 35.2

0.500 6.46 32.9

0.500 5.87 29.9

0.500 6.08 30.9

0.500 5.51 28.1 30.5

0.750 13.82 31.4

0.750 9072 2200

0.750 12.38 28.1

0.750 12.84 29.2 27.7
T/D = 0.500

 

Diameter (in.) Force (#2 F/A (#/in.2) Average F/A (# in.2)
0.1562 0.48 25.0
0.1562 0.53 27.6
0.1562 0.50 26.1 26.2
0.1875 0.57 20.4
0.1875 0.77 27.8
0.1875 0.79 28.5
0.1875 0.87 . 31.4 27.0
0.250 1.23 25.4
0.250 1.09 22.2
0.250 0.94 19.1
0.250 1.17 23.4 22.5
0.375 2.30 20.7
0.375 .1085 1607
0.375 2011 1900
0.375 2.28 20.3 19.2
0.500 3.88 19.8
0.500 3.53 18.0
0.500 3.28 16.7
0.500 3.35 17.0 17.9
0.750 6.62 15.0
0.750 6.44 14.6
0.750 6.42 14.5
0.750 5.87 13.3 14.4

 

 

 

61

TABLE IV
Adhesion of Polysulfide Elastomer to Steel
of Varying Surface Roughness
(Constant T/D Ratio)

A. Surface Roghness - 5.0 11in.

 

 

 

 

 

 

 

 

Diameter (in.) Force (#2 F/A (#/in.2) Average F/A (#/in.2)
0.375 1.82 16.4
0.375 2.01 18.2
0.375 1.32 11.9
0.375 1.37 12.4 14.7
0.750 5.27 11.9
0.750 6.89 15.1
0.750 6.31 14.3
0.750 4.91 11.1 13.1
B. Surface Roughness - 35.5 pin.
Diameter (in.) Force (#2 F/A (ff/in?) Average F/A (#/in.2)
0.375 3.82 34.4
0.375 4.02 36.2
0.375 3.29 29.6
0.375 4.51 40.6 35.2
0.750 13.82 31.4
0.750 9.72 22.0
0.750 12.38 28.1
0.750 12.84 29.2 27.7

C. Surface Roughness - 60.0 11in.

 

 

 

 

Diameter (in.) Force (#2 F/A (#/in.2) Average F/A (#/in.2)
0.375 4.60 41.4
0.375 4.85 43.7
0.375 5.24 47.3
0.375 5.47 49.3 45.4
0.750 12.44 28.2
0.750 12.56 28.5
0.750 14.17 32.2
0.750 12076 2809 2905

62

TABLE V

Adhesion of Polysulfide Elastomer to Glass
(Sample Configuration "B" - T/D = 0.125)

 

 

 

Diameter (in.) Force (#2 F/A (#/in.2) Average F/A (#/in.2)
0.243 4.16 85.5
0.243 2.61 56.3
0.243 2.93 63.2
0.243 3.13 67.5 68.1
0.375 7.20 64.8
0.375 6.43 57.8
0.375 6.30 56.7
00375 . 7070 6903 6202
0.735 12.44 29.3
0.735 12.98 30.6
0.735 12.52 29.5
0.735 13.07 30.8 30.1
TABLE VI

Adhesion of Polysulfide Elastomer to Ammonium Nitrate
(Sample Configuration "B" - T/D = 0.125)

 

 

 

Diameter (in.) Force (#2 F/A.(#[in.2) Average F/A (#/in.2)
0.605 2.34 8.18
0.640 1.26 3.92 ' 5.6
TABLE VII

Adhesion of Polysulfide Elastomer to Oxidized Steel
(Sample Configuration "B" - T/D = 0.125)

 

 

 

Diameter (in.) Force (#2 F/A.(#/in.2) Average F/A (#fin.2)
0.750 13.27 30.0
0.750 12.49 28.3
0.750 12.57 29.5
0.750 12.78 29.0 29.2
0.500 10.03 51.2
0.500 10.38 52.9
0.500 9.01 46.0 50.0

63

TABLE VII - (continued)

Diameter (in.) Force (#2 F/A (#/in.2) Average F/A (#/in.2)

0.375 6.67 60.0
0.375 5.62 50.5
0.375 8.07 72.7 61.1
0.250 2.91 59.4
0.250 3.10 63.2
0.250 2.28 47.4

0.250 3.70 76.2 61.6

APPENDIX B. ELASTOMER TECHNICAL DATA

ELASTOMER TECHNICAL DATA

PX-45 polysulfide base rubber is manufactured by the Permalastic
Products Company, Detroit, Michigan. The manufacturer's data sheet
states the following pr0perties of the cured product: resistance to
0115, solvents, mild acids and alkalies; resistance to oxidation,
ozone and weathering; moisture and gas impermeability; excellent low
temperature and dielectric pr0perties. The uncured polymer is stated
to have an indefinite shelf life. The consistency is that of a
pourable syrup.

The material as used in this work was cured using the PX+46 lead
base catalyst from the same manufacturer. The formulation used in all
cases was as follows: Polymer PX+45 - 20 grams, Catalyst PXA46 1 gram
and Toluene (to decrease viscosity) - 2 ml. These components were
carefully weighed and mixed by hand to a uniform.mixture. The average
pot life was 7 minutes with complete set in less than 30 minutes.

The set polymer was cured for four hours at 65°C. in all cases.

The chemical structure of the unconverted polymer is that of a

difunctional mercaptan. The average structure is as follows:

HS- (CeHu- 0- 0112- 0- CeHu- s-s ) 6—C2Hu-0-CH2- 0- C2H22- SH

Conversion of the unconverted polymer to a rubbery material is

accomplished by oxidation, addition, and/or condensation to form

sulfide crosslink bonds.

65

Y AXIS TRAVEL UN.)

 

66

 

 

O)?
\f/
9°“
0
\ \
‘I~ dag?
t \"
50"»
153°
1 I A A I L
2 3 . 4 5 6 7
9 I0 II l2 I3 I4

FORCE (LBS)
FORCE CONVERSION GRAPH

 

”IIIIIIIIIIIIIIIIIIIIIIIIIIIQllilllll'Es

1293 031