Reliability Prediction Studies On Electrical Insulation Navy Summary Report NAVAL

'T-77
72-
--
,IT777--
NRL Report 8095
Reliability Prediction Studies on 1

Electrical Insulation: Navy Summary Report
E. L. BRANCATO, L. M. JOHNSON, F. J. CAMPBELL
Naval Research Laboratory
and
H. P. WALIKER
Naval Ship EngineeringCenter
July 13, 1977
03
SEP'
15 1977
NAVAL RESEARCH LABORATORY

Washington, D.C.
Approved for public release; distribution unlimited.
SECURITY CLASSIFICATION OF THIS PAGE ("oen Data Entered)
READ INSTRUCTIONS
BEFORE COMPLETING FORM
REPORT DOCUMENTATION PAGE

I
REPORT NUMBER
2. GOVT ACCESSION NO.
3. RECIPIENT'S CATALOG NUMBER
NRL Report 8095

4. TITLE
~Final
(and Subtitle)
S.
TYPE
OF REPORT
PERIOD
COVERED
report
in the &NRL
Problem
6.
PERFORMING ORG. REPORT NUMBER
ON ELECTRICAL
RELIABILITY PREDICTION STUDIES
INSULATION: NAVY SUMMARY REPORT

7. AUTHOR(e)
S. CONTRACT OR GRANT NUMBER(&)
E. L. Brancato, L. M. Johnson, F. J. Campbell, and H. P. Walker

9. PERFORMING ORGANIZATION NAME AND ADDRESS
10. PROGRAM ELEMENT. PROJECT. TASK
Naval Research Laboratory

Washington, D.C. 20375
NRL Problem E02-04.801

SF 53-538.707
AREA & WORK UNIT NUMBERS
12. REPORT DATE
I1. CONTROLLING OFFICE NAME AND ADDRESS
July 13, 1977
13. NUMBER OF PAGES
I113
14.
MONITORING AGENCY NAME & ADDRESS(If diltc. .. t Itm Controlling Office)
15. SECURITY CLASS. (of Chfo repot)
Unclassified
1Sa. DECL ASSI FICAl ON/DOWN GRADIN G
SCHEDULE
16. DISTRIBUTION STATEMENT (of this Report)
Approved for public release; distribution unlimited.
17. DISTRIBUTION STATEMENT (of lhb.,

bstact nte.d In Block 20. If different from Report)
IS.
SUPPLEMENTARY NOTES
19. KEY WORDS (Continue on reverse side If necoeee'y and Identify by block numbs?)
Aerospace wires
Aging procedures
Combined environments
Cyclic temperature life
Electrical insulation
20.
Humidification
Life predictions
Magnet wire
Motorettes
Nonlinear data
Radiation effects
Regression analysis
Thermal classification
Thermal life
Truncated data
ABSTRACT (Contiue on reverse side If necessary and Identify by block number)
Reliability prediction studies on magnet wire, motor Insulation systems, and aerospace wires,
conducted over a span of 20 years, are summarized. Emphasis is on thermal aging of insulation, including
problems in developmehit of evaluation methods and equipment, analysis of data, and development of
Navy and industry standards. The usefulness of truncating data to save time is discussed, and truncation
is found to save up to 20% in testing time. New and improved equipment is described, including a novel
(Continued)
~FORM
DD
JAN 73
1473
EDITION OF I NOV 6- IS OBSOLETE

S/N 0102-014-6601
SECURITY CLASSIFICATION OF THIS PAGE (Wh.n Data fnflerd)
S44
la
.1
-,tL1JNITY CLASSIFICATION OF THIS PAOE(Whon Data E~ntered)
20. (Continued)
condensation chamber that provides a controlled wet humidity cycle during functional evaluation of insula.
tion systems. A recommendation to adjust the Navy's 40 000-h lifeline benchmark is made, on the basis
of recent data and field experience. Regression analysis data for 254 magnet wire twist tests and 62
insulation system motorette tests are documented.
11
SEUIYCASFCTOVF
TISPG
~t n
Dt nee)-
CONTENTS
Page
INTRODUCTION ...................................
ORIGIN OF THERMAL CLASSIFICATION AND

EVALUATION .....................................
PHILOSOPHY OF THERMAL CLASSIFICATION .........
THE NEED FOR THERMAL EVALUATION

PROCEDURES ........
............................
18
DEVELOPMENT OF EXPERIMENTAL EQUIPMENT ......
25
Magnet Wire Twist Test Procedure ..................

Motorette Test Procedures ........................
TREATMENT OF DATA .............................

Calculation for Placing Regression Line ...............
Extrapolation Problems with Nonlinear Life Curves .....
Truncating Data to Shorten Thermal Aging Tests ........
AEROSPACE WIRES ................................
Application of Thermal Evaluation Principles to
Aircraft W ires ................................
Cyclic Temperature-vs-Life Calculations ..............
Applications ...................................
FUTURE WORK ON PROBLEMS IN NEED OF SOLUTION
25
28
42
42
51
56
60
60
62
63
65
Humidification .................................
66
Failure Criteria .................................

Duration of Tests ...............................
Combined Environments ..........................
66
66
66
CONCLUSIONS AND RECOMMENDATIONS .............
67
ACKNOWLEDGMENTS ..............................
68
REFERENCES .....................................
68
Appendix A-Use of Computer Readout Tables Al and A2 ...
70
iii
LA
RELIABILITY PREDICTION STUDIES ON ELECTRICAL

INSULATION: NAVY SUMMARY REPORT
INTRODUCTION
The Navy's combat effectiveness depends on the reliability of the equipment in its
vessels. The reliability of electrical equipment can be increased by overdesigning and by
decreasing operating temperatures. This might be practical in land installations where
weight and size are unimportant. On a ship, the resulting increased weight and size would
decrease the combat effectiveness of the vessel. Thus, equipment designers must reduce
weight and bulk and also maintain and even improve reliability. This could be achieved
with the improved synthetic materials which were, eing generated by industry at a considerable rate, during the post-World War II period. While this generation of materials displayed superior characteristics, they had no pedigree to assure long life in use.
The Navy, faced with the responsibility of increasing the combat effectiveness of its
vessels, embarked on a major study of the aging properties of electrical insulating
materials and systems. This involved the development of evaluation techniques for comparing the expected lives of new insulations. This program was initiated at the Naval
Research Laboratory in 1952. Around 1965, when techniques for evaluating electrical
components and systems were being firmed, much of the evaluation program was transferred to the Navy Ships Research and Development Laboratory.
This research, while it benefited the Navy, also had considerable impact on national
and international standards. To illustrate a few highlights, the data generated by the
Navy confirmed the thesis that thermal aging was governed by the chemical laws known
as the Arrhenius relations. The influence of electrical and r.iechanical stresses and of
humidity were demonstrated. In this connection, NRL established the level of mechanical stress that has been universally adopted in the applicable standards. One of the
problems was the level of the lifeline at which temperature ratings could be compared for
various insultations. The Navy provided much of the long-term temperature-life data needed
for establishing this standard. Humidification, which was used as a sear-hing agent for failures, was difficult to standardize. However, when the IEEE 117 Test Procedure was being
developed the Navy took the initiative to investigate this factor and produced a technique
and a chamber design that were incorporated in the procedure.
Over the years, all thermal aging data were universally treated as an Arrhenius
relationship. However, the question arose, in connection with aircraft wire aging, of
whether life predictions could be made when the operating temperatures were variable.
This was demonstrated to be possible by integraing the effects dictated by the Arrhenius
laws. In addition, a significant contribution was made to the economics of the thermal
Manuscript submitted December 9,1976.
/:
.. .. .....
i;
....
BRANCATO, JOHNSON, CAMPBELL AND WALKER

evaluation procedures. In general, a year of experimental time is required to obtain a
life-vs-temperature characteristic curve. A detailed statistical- analysis validated truncated

data techniques as applied to thermal aging; this substantially reduced testing time.
ORIGIN OF THERMAL CLASSIFICATION AND EVALUATION

SThe
need to know the effects of temperature on electrical insulation was first

analyzed and discussed when Steinmetz mid Lamme published their 1913 paper,
"Temperature and Electrical Insulation" [1]. This classic paper not only reflects the
accepted concepts of their day but also introduces the theory that insulation deteriorates
with time at certain temperatures.
At that time insulating materials were classified in three main categories, known as
classes A, B, and C, according to the general compositions of the materials. Class A included fibrous materials such as paper and cotton, along with most of the natural oil
resins and gums. Class B included heat-resistant materials like mica, asbestos, and
equivalent refractory materials, frequently used in combination with other binding
materials. The fireproof or heatproof materials such as mica, "so assembled that very
high temperatures do not produce rapid deteriorr' )n," were considered Class C.
In 1913 the accepted general temperature ranges for the three classes were
Class A - to 90 0 C
Class B - to 1250 C
Class C - to 150 0 C to point of incandescence.
Figure 1 is a general guide temperature-life curve for Class A insulation taken from
the 1913 Steinmetz and Lamme paper. It illustrates the generally prevailing belief that
electrical insulation suffered insignificant deterioration below 90 0 C but that above 100 0 C
the rate of deterioration increased rapidly until 125"C, above which life was shortened to
a few weeks. In other words, it was thought that aging did not begin until a definite
temperature had been exceeded.
It is interesting that it was then also thought that if the insulation was cooled to
room temperature between duty heat cycles, the actual hours of accumulated thermal
aging would be decreased (as compared to continuous duty) because the insulation would
have a chance to "recover."
By the late 1920s a higher figure, 1050 C, was taking hold as the representative
temperature for Class A materials, although V. M. Montsinger in his 1930 paper, "Loading
of Transformers by Temperature" [2], advocated a more conservative value of 95 0 C. In
addition, Montsinger believed that the sole end-of-life criterion was mechanical failure of
the insulation and that it was "hopeless to judge the rate of deterioration of insulation
by its electrical strength." This idea stemmed from the belief that the electrical strength
of insulation increased in general with age until the material actually cracked open. At
the same time he introduced the idea that mechanical deterioration was a continuous
reaction to temperature and that the rates could by some means be determined. This
was in sharp contrast with Steinmetz, who held that no deterioration could occur below
a critical temperature for the material.
-2
NRL REPORT 8095
6-
--
80
100
120
140
TEMPERATURE (DEGREES
F1
160
180
200
C)
Fig. 1-Possible life-vs-temperature relationship of

Class A insulation [1]
The accumulated data from his study suggested a general law for insulation aging
that is represented by a straight line on semilog paper with a linear temperature scale.
The curve was expressed by the equation
J0
Y = Ae-mt
where Y = life in years
A and m are constants that characterize the insulation
t = temperature in C
e = base of natural logarithm.
This data, obtained over nine years, was on the tensile strength of paper, aged in oil
and in air. A product of this milestone study of Montsinger's was a rough demonstration
of what might have been called an "8- or 10-degree" rule. Additional data obtained during
the subsequent years substantiated this rule, defining it as the "10-degree" rule. In effect,
this empirical relation states that the thermal life of insulation is halved for each 100 C
increase or, conversely, doubled for each 100C decrease.
31
--:.-
-7

In the development of functional evaluation there has been and probably always will
be one primary question: Have all the factors that produce measurable and significant
changes in the life been properly included? Naturally, the purest form of functional
evaluation would be the gathering of life data on the equipment itself when used under
specific field conditions. However, this is impractical for two good reasons, namely time
and expense. Yet a flood of new materials were becoming available, starting in the
middle 1940s and increasing in volume through the 1950s.
Industry as well as the government realized that it was imperative to devise some
means to evaluate functionally insulating materials and even insulation systems relatively
quickly and at reasonable cost.
Thus, a new milestone had been reached as numerous laboratories engaged actively
in a comprehensive program, embracing many approaches to the problem and eventually
resulting in new test procedures. The Navy in particular had a compelling desire to move
ahead in materials and insulation systems engineering, because of its urgent need for
military specifications for purchasing these new materials. The first research to be performed was to investigate the effects of voltage, vibration, heat cycling, and humidity on
the life of magnet wire insulation [3]. To this end, coils were wound with about 30.5 m
(100 ft), of number 26 enameled magnet wire. Ten turns of wire of the same insulation
were inserted at the center of the coil to provide (a) a known dielectric stress, (b) detec-
tion of insulation failure by current flow, and (c) measurement of temperature by

resistance change. The coils were heated by circulating current through the main winding,
in a constant ambient of 600C. The assembled coils on the mounting rack are illustrated
in Fig. 2.
q\0
Fig. 2-Coil assembly showing asbestos wrapping and lead attachment
II
NRL REPORT 8095
With the above experimental setup, the effect of voltage at 1 g, from 4.5 to 50 V,
was first investigated. As in Fig. 3, there was a strong dependence of life on voltage at
1600C. The effects of vibration were determined at 35 V d.c., from 0 to 5 g at 160*C.
Figure 4 shows the effect of vibration to be more dominant at higher values of g, but
the spread in life data is much less.
During these early investigations by the Navy, industry was conducting parallel research to develop evaluation techniques. However, they were concentrating more on the
development of test models to simulate magnet wire applications and complete insulation
systems for electrical rotating machinery.
By 1952, Dexter was experimenting with modifications of the standard NEMA
magnet wire twist sample formerly used for voltage breakdown testing. A generous
amount of thermal aging data were already being accumulated using various test variables
including temperature and voltage stress, which he reported in 1954 [4]. This together
with the work of many others, including the members of AIEE committees, constituted
the background experience that later made such classic documents as AIEE 57, 65, 510,
and 511 possible.
Simultaneously, Cypher and Harrington were developing a test model of a motor
that would be suitable for evaluating functionally an insulation system without the expense of full-size motors [5]. Later these model motors were appropriately named
"motorettes." This marked another milestone in the development of functional evaluation, as ,it paved the way to several other "model-ettes," among them the armette and
formette. For several years the motorette went through development and design improvements but remained basically the same. Two of these improvements were rather significant. First, in a study by the Navy, it was found that the Class H terminal block was
allowing excessive current leakage under high humidity; this was solved by substituting
porcelain and standoff insulators. Second, an almost two-to-one savings of space and
weight were realized when John Dexter's smaller motorette design was adopted; the current design is shown in Fig. 5.
During this time physic's and chemists such as Dr. Dakin were making a closer
study of the basic phenomena of thermal aging of electrical insulation. In 1948, he
published a paper [6] proposing a chemical rate theory interpretation of thermal deterioration. This was a logical proposal since the observed physical changes during the
thermal aging are the results of internal chemical change. It not only provided a more
satisfactory explanation, but also allowed a more correct coefficient of deterioration
than was permitted by the 10-degree rule. This more descriptive relationship is
L
AeB/T
where T is the absolute temperature and A and B are constants determined by the activation energy of the particular reaction (or log L = logA + BIT). Thus, plotting the log of
the life of the insulation against the reciprocal of the absolute temperature should produce a straight line. This relationship was generally confirmed except where second and
higher order chemical reactions enter into the aging phenomena.
-A
100
600-
~~500
aI'
Fig. 3-fetof
300
200-
voltage
onlieo
Formex insulated coils at 1600 C,

1.0-g vibration. Lines a-b represent
95% confidence intervalA
10000
10
30 40
20
(g)
VIBRATION
50
1000
900-
NO0 b
PFrepresent
Fig. 4-Effect of vibration on life of

Formex insulated coils at 160 0 C,
35 V d.c. dielectric stress. Lines a-b
95% confidence interval
Z;~
500400-
300
20 30 40L 50
010,
00
STRESS
(VOLTS
OIELECTM~
zi~
NRL REPORT 8095
v,
4o
'
Fig. 5-Assembled motorette specimen with

varnish treatment
PHILOSOPHY OF THERMAL CLASSIFICATION

During the past two decades much experimental work has been done to investigate
the concept of temperature classification of magnet wire insulation, based on the possible
linear relationship between the logarithm of life and the reciprocal of absolute temperature, as first observed by Dakin [6] to follow the Arrhenius chemical deterioration rate
equation. Before this, the only available method was to assign temperature classifications
based solely on the types of materials. This obviously left much to be desired.
The IEEE and ASTM, recognizing the significance and value of an evaluation method
that was suitable for use in the laboratory under accelerated test conditions and yielded
extrapolated data suitable for field classification purposes, sponsored development of the
necessary test procedures. However, any accelerated laboratory test procedure for
evaluating life-temperature characteristics of insulation can produce no more than comparative values on various insulations. Because of this limitation, one of the basic requirements is to establish hours-of-life reference values necessary to convert the laboratory
life values of the various insulations into temperature ratings for field use. For example,
insulation "A" has been proved through field experience to have earned a temperature
rating of 1050C for a normal life expectancy of 15 to 20 years. This same insulation
under laboratory test yields an extrapolated life of 5 years at 1050C. Based on this
information, a newly developed insulation "B", which also yields 5 years of extrapolated
life at 1050C, would also qualify for the same temperature rating. In like manner, if
insulation "C" yields 5 years of extrapolated life at 130 C, it would qualify for a 130 0 C
temperature rating.
7
tq
Rn 'Government
It is only logical, then, that this reference life value should evolve from an insulation
for which there is already adequate field experience. In addition to the field experience,
adequate laboratory data must be available on the same insulation. Both industrial and
laboratories have, over the past 20 years, accumulated much data on polyvinyl formal coated magnet wire, using IEEE and ASTM test procedures [7]. This
wealth of laboratory data confirm actual field experience on the same wire and provide a
basis for establishing reference lifelines for realistic temperature classification of filmcoated magnet wires.
Most field experience with polyvinyl formal magnet wire has been with equipment
impregnated with phenolic varnishes. This experience has firmly established a 1050C
temperature classification for this wire. On the other hand, laboratory tests have indicated that at 1050C polyvinyl formal enameled wire, impregnated with the same varnishes,
yields an average extrapolated life of 40 000 h (approximately 5 years). This value is based
on many governmental and industrial laboratory tests with test temperature points ranging
from 1300C to 2000C. It is recognized that due to many variables throughout the various
laboratories, a wide range of extrapolated life values does exist. This range has a spread
of from 30 000 h to as high as 90 000 h, depending on many such factors as the faithfulness to the test procedure, the test temperature range, and the particular phenolic
varnish used. However, enough data were on hand at the time to set a minimum average
life value of 40 000 h. The Naval Research Laboratory and Navy Ships Research and
Development Laboratory have investigated well over 200 film and varnish combinations
in the past 20 years, accumulating life-temperature data in many cases of well over 10 000
h at the lowest test temperature. Figures 6, 7, and 8 illustrate some of the NRL investigations that support the 40 000-h life figure for polyvinyl formal magnet wire impregnated
F_
with phenolic varnishes.
It is generally recognized that some manufacturers find the need to rate their filmcoated magnet wires thermally, based on unvarnished temperature-life data. To make
this kind of classification requires that an equivalent standard be established for unvarnished film-coated wires. The field-proven "benchmark" magnet wire (polyvinyl
formal) yields an extrapolated life of 20 000 h at 1050C when tested unvarnished under
the referenced test procedures, as illustrated in Fig. 9. Again, this figure was established
by both governmental and industrial laboratory test and offers a basis for establishing
a reference classification lifeline equivalent to the 40 000-h lifeline for varnished wires.
If equal thermal life could be expected from varnished and unvarnished film-coated
wires, one common lifeline could be used for classification. However, experience has
shown that this is not always the case. In a mimber of cases vainished wire yields a
two-to-one, or better, life over unvarnished wire. See Figs. 9, 10, and 11 for typical
examples. This two-to-one difference is significant in the case of polyvinyl formal coated
wire (Fig. 9) inasmuch as it has been the oie magnet wire (used in varnish-impregnated
electrical equipment) with enough years of field experience to earn a rating of 1050C.
Therefore the Navy, for one, has expected all new magnet wire coatings to meet the
laboratory test benchmark reference lifeline of 40 000 h at the claimed temperature
rating.
NRL REPDORTI
8095
zon-
100
\J
IIVES AND 95% LOWER ONLY CONFI-
EY
0
z
90
100
110
120
130 140 150
TEMPERATURE,
160
170 180 190
210
230
C (1/K SCALE)
Fig. 6-Life-temperature characteristics of twist combination no. 9, polyvinyl

formal magnet wire impregnated with phenolic-type varnish
175E
R7 100
011
FA
0I
3A-
.17
90
100
110
120
130
140
150
TEMPERATURE,
160
170 180 190
210
C 11/X SCALE)
Fig 7-Life-temperature characteristics of twist combination no. 10,

polyvinyl formal magnet wire with phenolic-type varnish
10
230
ni r --- -i
,-
_ __
__ ,-_
NRL REPORT 8095
0_
so
kV
z
0
90
100
110
120
130 140 150 160
TEMPERATURE,
170 130 190
210
C 11/K SCALE)
Fig. 8-Life-temperature characteristics of twist combination 135F,

polyvinyl formal magnet wire with phenolic-type varnish
I1
230
100
0-VARNISHED WITH PHENOLIC-TYPE

COMB. NO.
(TIS
r\
0
10
-UNVARNISHED
0.-
90_
100
110
120
130
140
150
160
(TWIST COMB. NO. 8)

-
170 180 190
210
TEMPERATURE. CC (1/K SCALE)
Fig. 9-Life-temperature characteristics of varnished and unvarnished

polyvinyl formal magnet wire
12
230
NRL REPORT 8095
VARNISHED (TWIST CM.N.81)
bc0
___
UNVARNISHED (TWIST COMB. NO. 80
\-
Lc
IX-
90
100
110
120
130
140 150
160
170 130 190
210
230
TEMPERATURE, *C Il/K SCALE)
Fig, 1O-Life-temperature characteristics of unvarnished and phenolicvarnished epoxy-overcoated magnet wire
ZII
0c-VARNISHED(TITCM.N.2)2
\A-UNVARNISHED (TWIST COMB. NO. 24)_-
S-
50
--
120
130
A---
90_
100
110
140 150
TEMPERATURE,
160
170 160190
210
230
OC l1/K SCALE)
Fig. 11-Life-temperature characteristics of unvarnished and phenolicvarnished polyester-overcoated magnet wire
144
-4
NRL REPORT 8095
Rpolyvinyl
k,
Vcoat)
Many of the newer polyester-type magnet wires either yielded the same thermal life
or were downgraded when treated with an impregnating varnish. Because of this, opinion
was that the new film-coated magnet wires should be thermally rated using a common
varnished-unvarnished lifeline of 20 000 h which had been established by unvarnished
formal magnet wire at 1050C. All current evidence dictated a 40 000-h classification lifeline for rating film-coated magnet wires to be used in varnish impregnated
rotating electrical equipment. The issue was debated for several years in an attempt to
justify the varnished-unvarnished reference standard of 20 000 t.. As newer, higher
temperature magnet wires appeared on the market the Navy continued to rate them on
the basis of the varnished polyvinyl formal field experience benchmark. In the meantime,
however, industry was recommending a rating of about 10 0 C higher to their buyers than
would be justified on the basis of the polyvinyl formal field experience benchmark.
After 15 years' use of the now common modified polyester (with and without top
magnet wire for use at 1800C, the evidence now points to a "satisfactory" field
experience record at this operating temperature. A "satisfactory" record has been defined as one that does not include an undue number of negative reports or complaints
against the product when used as recommended by the supplier. If the polyesterovercoated magnet wire is considered on the same basis as polyvinyl formal (as far as
having earned a satisfactory field experience record is concerned), it likewise becomes a
benchmark candidate. Fortunately, the Navy has adequate laboratory life-temperature
data to set the lines for varnished as well as unvarnished tests. The data indicate that
aging tests were conducted at low enough test temperatures to produce lives well above
10 000 h, and even 15 000 h in some cases. The state of the art in recent years has produced even more reliable %.gingdata than was available when polyvinyl formal magnet
wire was being laboratory ',valuated for the purpose of setting benchmark values. When
the laboratory test data for the modified polyester-overcoated magnet wire are plotted
(Fig. 12), it can be seen that if the 1800C extrapolated operating temperature is considered, the reference lifeline becomes 20 000 h for the magnet wire when varnished with
typical polyester varnishes used to impregnate rotating electrical equipment.
When the combined laboratory test data for the motorette systems employing the
same modified polyester-overcoated magnet wire and class 155 varnishes are plotted
(Fig. 13) a lifeline of 20 000 h yields 1840C, as compared to 182 0 C for the varnished
magnet wire twist tests. This 20C higher extrapolated temperature for the motorette
tests provides a somewhat more conservative estimate for selecting 180*C as the appropriate qualifying temperature at 20 000 h. In considering this correlation between twist
and moterette test data it should be noted that the motorette data were obtained using
the Navy's version of the IEEE 117 Test Procedure which employs a humidity cycle of
100% relative humidity with no visible condensation. (Approximately half the life can
P"
be expected if 100% relative humidity with visible condensation is used, as stipulated in

the IEEE 117 procedure.) It is also important to note that the motorette insulation systems tested were "supported" systems. This means that the phase and ground insulations
support the magnet wire, allowing turn-to-turn first failures to reflect the life of the
magnet wire component in the system.
15
152-S4
100
7so-
MEAN REGRESSION LINE AND 95% L.C.L. ONLY

COM91INED DATA OF 13 TWIST LIFE TESTS
(TWIST CCMBINATIONS 69,82,100,161,196,199,273,274,275,
276,277,278 AND 323)
AVERAGE OF ALL DATA AT GIVEN

TETTEMPERATURE.
\\-LOG
10
--
--
~-
Lz
0
I..
V
TEMP. INDEX AT 20000 h-1820C
TEMP. INDEX AT 30000 h-17 0C

TEMP. INDEX AT 40000 h-172 0C
140
130
160
170
180
190 200 210 220 230 240
TEMPERtATURE,
260
280
*C f1/C SCALE)
Fig. 12--Life-temperature characteristics from combined aging data for 13

twist combinations using polyester-overcoated magnet wire and Class 15
impregnating varnishA1
16
NRL REPORT 8095
1oc0
-
MEAN REGRESSION LINE AND 96% L.C.L. ONLY
COMBINED DATA OF 7 MOTORETTE TESTS

(SYSTEMS 58,60,75,77,78, AND 81)
AVERAGE OFALDTAT
_x-LOG
GIEN.TSTTEMPERATURE.
44
INDEX AT 20000 h-1840C-TEMP. INDEX AT30000 h-1800C

-TEMP. INDEX AT 400M0h-177 0 C
-TEMP.
140
150
160 170
180
190 200 210
TEMPERATURE,
--
220 230 240
260
250
C (1/K SCAME
Fig. 13-Life-temperature characteristics from combined aging data for

seven motorette systems using polyester-overcoated magnet wire and Clacs
155 impregnating varnish
17
When thermal aging tests are conducted on the same magnet wire unvarnished, the
temperature index moves up from 180'C to 186*C at 20 000 h, as shown in Fig. 14.
This higher rating for the unvarnished magnet wire reflects a thermal downgrading due to
the varnish. Fortunately, this is offset by the mechanical bonding and the sealing out of
contaminants, which often are critical factors. It has been found that
certain higher
temperature varnishes, such as the silicone-modified ones, do not downgrade the thermal
life but instead upgrade it by as much as 12 0 C, as illustrated in Fig. 15 and 16. Unfortunately, these varnishes have a considerably lower bond strength characteristic, which
renders them unsatisfactory for many applications.
When the newly established polyester magnet wire benchmark of 20 000 h is applied
to the varnished polyvinyl formal magnet wire thermal aging data, it shifts the thermal
index rating from 105CC (at 40 000 h) to 1150C (Figs. 17 and 18). The 1050C temperature index was more or less arbitrarily chosen for the polyvinyl formal before field experience dictated it. Therefore, the higher figure of 1150C may be an appropriate rating
that would have been justified in the same manner as the 1800C rating for the polyesterovercoated magnet wire. In fact, there is now evidence that the 1050C polyvinyl formal
rating may be a conservative figure. European motor manufacturers have been rating polyvinyl formal well above 1050C (at 120 0 C) for many years [8], and in this country the
transformer industry [9] has done likewise.
THE NEED FOR THERMAL EVALUATION PROCEDURES

As a purchaser and user of insulating materials, the Navy naturally has a keen interest
in both the development and the end use of functional test procedures. The Navy
that proper development and use of these procedures would greatly benefi both
industry and military.
Srealized
The purpose of thermal endurance tests may be divided into two general categories,
as follow s:
zt
1en.To aid in selection and procurement of insulating materials for electrical equipSment
that will achieve maximum reliability at minimum cost
Ao
2. To provide engineering data that will ensure the fullest use of the potentials of
these materials when used in combinations, as systems, in electrical equipment.

In reference to the first purpose, the Navy must use the best available means of
screening insulating materials for military purchase and use. One of these means is
temperature classification of materials based on thermal evaluation tests. Since there are
existing materials that have been accepted for certain temperature classifications based on
long-time field experience, the life-temperature characteristics of these materials determined
by test provide a basis for comparison with the thermal life of new materials. The reason
for assigning these materials to definite temperature classes is to provide this means of
comparison and to designate each class by a single number for purposes of standardization.
To accomplish this, the Navy must set concite and definite limits based on careful consideration of the many factors involved.
18
j:g
NRL REPORT 8095
ILE
RERSSO
LINE AND %.;..!N1Y
COMBNED
ATAOF 5 TWIST LIFE TESTS
(TITCOMBINATIONS 68,98,272,321 AND 322)
so-~
*~X TEMP.
INDEAG
AT 2000
TEMP.~
INDXMAPERATURESC
~
ES
T!
Fig
PR
T R,
chrceitc
unanse0ws
et
DATA ATGIVE-
3j
(/
CL)-
A4Lf-eprtr
Irmcmie
fpletroecae
gn
aafrfv
antwr
:E3
TEM.
IDE AT000 h196C
TEMP
81 INDX AT3000 h-
100
so'\
o0
0,
--
SAE
10
B7
I--
kii
--
02
140
150 160
170
150
190 200 210 220 230 240
TIMPERATURIo C
(1/
260
210
K SCALE)
Fig. 15-Life-temperature regression line and log average lives for polyesterovercoated magnet wire twists varnished with modified silicone varnish
(twist combination no. 70)
20
NRL REPORT 8096
I-o
bc
0 o
a3
.5
140
150
160
170 160
190 2002110 220 230 240
26-
TEMPERATURE, *C (1/DK SCALE)
Fig. 16-Life-temperature regression line and log average lives for polyesterovercoated magnet wire twists varnished with modified silicone varnish
(twist combination no. 99)
21
96% CONFIDENCE LIMITS
.0
0101010101010
'A
-,--~-~-
-~
--
NRL REPORT 8095
REGRESSION LINE OF LOG

AVERAGE LIVES AND 96%
LOWER ONLY CONF. LIMIT
90
10
10
10
10
PR
-1/KSAE
TE
1010
TRE
Fi.1Iletmeaue hrceitc ftis obnto o 3B

poyiy omlmge iewt hnlctp ans
K2
.1J
-o
As for the second purpose of test procedures (to provide engineering data), the
Navy recognizes that the thermal endurance characteristics of insulating materials may not
correlate with the performance of those materials when combined in systems. This was
recognized as much as 15 years ago, when an article was published reporting the thermal
investigations of 13 insulating systems functionally evaluated by use of the motorette
[10]. Following are quotations from this article:
1. "It has become apparent that the thermal stability of various components is influenced by the aging characteristics of the companion components of the system. The
useful life of organic varnished glass phase material can be increased as much as 900% by
the proper selection of the magnet wire."
2. "The interrelationship of the individual components and their influence on the
first failures of a system is a critical factor in determining the aging characteristics of the
system."
This important interrelationship was dramatically illustrated several years later during
the IEEE 117 round robin test program, when a neoprene-treated tie cord was used in the
fabrication of the motorette test specimens [11]. The incompatibility was a reaction
between the neoprene and the polyvinyl formal magnet wire, causing erratic and premature
wire failures to occur under the tie cord.
It is recognized that a wide variety of results can be obtained, depending on such
factors as the test procedure followed, the faithfulness by which the procedure is carried
out, and the various test conditions employed.
Thermal evaluation data can provide a good indication of minimum performance
requirements, whether it be on materials for the purpose of screening and purchasing or
on complete systems for gaining engineering design data for equipment specifications.
Military Specification MIL-E-917D (Navy), covering the basic requirements of electrical power equipment for naval shipboard use, explains clearly the differences between
materials classification and systems classification. Paragraph 3A.1.10, in particular, says
that "a material that is classified as suitable for a given temperature may be found suitable
for a different temperature, either higher or lower, by an insulation sy8tem test
procedure." It is further pointed out in par. 3U .2 that "experience has shown that the
thermal life characteristics of composite insulation systems cannot be reliably inferred
solely from information concerning component materials."
Navy experience has shown that materials testing of such components as film-coated
magnet wire and impregnating varnishes provides a better basis for temperature classification than other supporting materials because the performance requirements are not too
different in most applications. This was supported by the report on the 13 systems
mentioned in Ref. 10. In fact, it was shown that where the phase and ground (slot) insulation supported the magnet wire (allowing the magnet wire to fail first), there was
reasonably good correlation between the wire insulation life of the system and that
obtained by the ASTM D2307 Twist Test. It should be pointed out here that the
motorette procedure used was the Navy's version of the IEEE 117 procedure, employing
a highly controlled humidity cycle using 100% relative humidity without visible condensation.
24
M7
NRL REPORT 8095
Because of the consistency of aging data that can be obtained by careful adherence
to the ASTM D2307 procedure and the modified IEEE 117 procedure, it was determined
that material classification by temperature classes was feasible for purchase specification
purposes such as those outlined in the J-W-00117 specification [12]. It must be remembered, however, that this was done only after rigid rules on use and interpretation of the
data were specified. For example, extrapolation is allowed only after realistic and appropriate requirements have been met in regard to linearity and other factors such as
maximum and minimum average life data.
DEVELOPMENT OF EXPERIMENTAL EQUIPMENT
Magnet Wire Twist Test Procedure
During the early stages of development and use of the twisted pair magnet wire procedure (IEEE 57) the method required handling each individual specimen throughout the
prescribed heat aging and voltage stressing cycles. Thus, for a temperature-life data experiment .with four temperature points a minimum of 40 specimens presented a problem of
handling and experimental processing.
Two variables (both mechanical) were unavoidably introduced by the individual
handling of the specimens as they were removed from the container (which was in the
form of a box, tray, or wire basket), placed on the voltage stressing apparatus, and later
returned to the container. The first was possible damage due to the handling itself, and
the second was adhesion damage experienced with some magnet wires when specimens
were placed together or on top of each other in the containers. Adhesion of the wires
was due to the plastic flow of the wire coating during the earlier stages of thermal aging.
This resulted in actual rupture of the insulating film as the specimens were pulled apart
at room temperature for voltage stressing.
As a solution to these problems the multiple twist specimen holder, shown in Fig.
19, was developed at NRL. The holding fixture permits mounting the specimens in
fixed, permanent positions, thus protecting them from these damaging conditions. The
assembly can be handled as one unit, eliminating the need of a container or tray. It
offers other advantages as well. The open sides of the frame provide more adequate circulation of air than a box or similar container. Also, the time required to voltage stress
the specimens is reduced to one-tenth of the conventional time required if the stressing is
done with a multiple tester such as the one designed at NRL for use with the holder (see
Fig. 20). A study conducted at the Naval Ships Research and Development Laboratory
[13] comparing several variations of the NRL twist specimen holder shows little if any
significant difference in the results obtained.
An additional contribution was made to the magnet wire twisted pair test procedure
in the refinement of the original specimen forming jig. The NRL forming jig, illustrated
in Fig. 21, allows for a more uniform and faster method of making twist specimens. It
was discovered that the angle at which the wire was formed as it left the twisted portion
of the specimen was a variable contributing to premature failure of the insulation. The
use of the forming jig not only drastically reduces fabrication time but also almost entirely
eliminates handling of the specimen during the process.
25
AA
r L
Ai4
Fig. 1 9-NRL multiple-twist-specimen holder
26
-~-~~
Am
NRL REPORT 8095
Fig. 20-NRL multiple-twist specimen voltage-stress tester]
Fig. 21-NRL twist specimen forming jig
27
One major limitation to the magnet wire twisted pair test procedure was the original
twist-forming apparatus, which worked satisfactorily only with film-coated wire. When
attempting to make fibrous-covered magnet wire twisted pairs, the rough covering of the
two adjacent wires would bind and rupture the insulation. Because of this problem
thermal aging of fibrous-covered magnet wire has not been required by government and
industry specifications.
I:
The solution to this problem was to design a special universal twist-making device
for use with fibrous as well as film-insulated wire (Fig. 22). In use of this device, a loop
of wire is suspended vertically with individual weights attached at the two free ends,
guided by two plastic pulleys. Unlike the original device the weights rotate freely, allowing twists to be formed without any grabbing or binding of the wire.
Thermal aging tests were conducted on film-insulated magnet wire twisted pairs
made using both the original and the new type of device to compare the resulting aging
data. The tests indicated' that the thernal life was the same regardless oi which device
was used. Thus, it was concluded that the device was suitable fcr making twisted pairs
from either type of insulated magnet wire.
Motorette Test Procedures
Although the IEEE 57 (later adopted by ASTM as D2307) magnet wire test procedure proved to be a significant contribution to screening and classifying magnet wire
according to temperature classes, it is limited inasmuch as it is no more than a materials
or component evaluation. Except for magnet wire-varnish combinations, it does not take
into consideration the interrelationship of other materials and the functional life reflected
in a complete insulation system. The IEEE 117 motorette procedure [14] was developed to
meet the need for a more functional systems evaluation procedure. The motorette models
the elements of a randomly wound motor. Its components consist of two bifflar wound
magnet wire coils so that conductor-to-conductor electrical tests can be made. The two
coils are inserted in a slot section and are insulated from each other by sheet phase material
and from the motorette frame by slot liners. Slot wedges are placed in the slot, compressing
the coils together to reduce coil motion. Details of parts and assembly may be found in
the IEEE 117 Test Procedure. Because of the unavoidably complex nature of the
motorette procedure, it involved many more variables and consequently presented many
experimental problems that had to be solved before it could be considered reliable and
useful.
Standardization
After the original round robin motorette test program was conducted to check correlation between laboratories using the IEEE 117 Test Procedure, it became apparent that
the test results were not very closely in agreement. The reasons for poor correlation
were difficult to determine because there was no certainty as to the identity of materials
used or the degree of faithfulness to the procedure. Yet it was necessary to determine
the limitations of the procedure's accuracy if the relatively high cost of the procedure
was to be justified. The Naval Research Laboratory contributed to the investigation of
the variables by conducting an analysis of the motorette specimens used by each
28
Zz-'
---
NRL REPORT 8095
RE-I
IFI
Fig. 22-NRL universal twist making device
29
laboratory in the round robin [15]. One motorette from each of the nine participating
laboratories was obtained for thorough examination, to determine significant differences
as well as any fundamentally poor construction that might contribute to the lack of correlation.
The nine motorette specimens examined are illustrated in Fig. 23. Figures 24, 25,
and 26 show an enlarged cross section of the slot portion of each motorette. The small
motorettes "D" and "E" were submitted by laboratories that did not participate in the
round robin. Due to a move of its laboratory and a change in its activities, one participant in the motorette tests was unable to submit a sample. It is to be noted that these
samples were submitted 2 years after the round robin tests began; and in some cases
the motorette was manufactured from whatever materials were on hand and most closely
simulated the original specified components. Because of this, one should place more
emphasis on the placing of materials and other structural details than on the particular
materials employed. Some of the more significant variations included:
Wire:
14 to 19 turns; loose to very tight pack in slot.
Varnish:
0.05-0.25 mm (0.002-0.010 in.) build; very light to very dark; poor

to very good penetration into slot.
Phase:
Very loose to tight fit to slot liner; some "folded in," others "notched"
to fit slot.
Slot liner:
0- to 9.5-mm (0-3/8 in.) protrusion from slot.
Sleeving:
None, organic varnished glass, silicone varnished glass, vinyl over glass;
some sleeving placed in slot.
From the results of the analysis, it was concluded that significant assembly and
manufacturing differences do exist and may contribute to poor correlation among laboratories.
Although the motorette was designed for functional evaluation, it must be kept in
mind that it is subjected to far more severe environmental conditions in an accelerated
test program than an equivalent motor insulation system would ever experience in normal
use. Hence, even small deviations in construction or departures from standard procedures
aie amplified in their effects. Personnel directly associated with the manufacture of the
motorettes must be aware of the more precise techniques required as compared with
those of the usual production line product.
For a number of years the Navy has conducted a continuing research program,
sponsored by the Navy Ships Engineering Center, on the various parameters influencing
the thermal aging properties of electrical insulation systems. As the ramifications of this
study were so numerous, NRL invited the Naval Ships Research and Development Laboratory (NSRDL) to share in the research work. In the interest of maintaining uniform
accuracy in the joint experimental work, it was decided that a series of comparison
studies would be conducted to determine the degree of correlation between two sets of
life-temperature data, obtained at NRL and NSRDL, on a Navy standard insulation system.
30
0'
,z
~All
NRL REPORT 80951
Iu
EIII~IiE
Il
01
I
U
322
U0
ww
NRL REPORT 8095
.20
3
5.a
t0
330
BRANCATO, JOHNSON, CAMPBELL ANI) WALKER
F
I
c~.
0
0
Cu
'04,
~
4,
0
0
2
0
4,
o
UC,
Cu
CuCu
0
U
0
CS,
34
,~.'-.4.
-~..-...
NRL REPORT 8095
A typical Navy Class 105 insulation system was used for the motorette specimens.
The system consisted of polyvinyl formal magnet wire; organic varnished glass phase; and
organic varnished mica-glass ground materials. The system was impregnated with an oiimodified phenolic varnish. Identical materials were used by the two laboratories, and
each fabricated its own specimens. The IEEE 117 Test Procedure was followed except for
modification in the moisture cycle. The Navy version of the procedure used 100% relative humidity with no visible condensation in place of the 100% relative humidity with
visible condensation called for in the IEEE 117 procedure. It had been determined
several years earlier that no practical method was available that assured uniform and consistent visible condensation. (The Navy was currently developing a humidity cabinet in
an attempt to solve this problem.)
VE
RE
Both laboratories, cognizant of the susceptibility of errors due to the many variables,
established stringent control over the parameters. The significance of this control is
evident in the high degree of data correlation between the two laboratories. The NRL
data are plotted in Fig. 27, and the NSRDL data in Fig. 28.
WHumidity
ConditioningCycle
k1
Eon
During development of the motorette procedure the Working Group concentrated

one main endeavor, which was to seek out and correct various factors influencing the
test results. Probably the most critical factor with respect to both reproducibility within
one laboratory and correlation between laboratories is the humidity conditioning cycle.
The IEEE 117 procedure stipulates that "each specimen is to be exposed for at least 48
hours to an atmosphere of 100% relative humidity with visible condensation on the
winding." Experience has shown that to maintain this condition consistently throughout
the conventional "plus-dew" chamber is extremely difficult. An exchange of experiences
among various laboratories revealed that due to such factors as loading, chamber design,
and room ambient variations there has been a wide variance in humidity conditions,
ranging from heavy condensation to no visible condensation on some specimens.
In an attempt to find a practical solution to this problem, the IEEE 117 Working
Group asked the Navy to devise a chamber that would meet the following requirements:
1. Provide a uniform and consistent visible condenstaion throughout the test area of
chamber
Sthe
2. Perform independently of room ambient fluctuations

3. Be completely self-contained so as to not require external plumbing or wiring
-
4. Accommodate up to 60 motorette specimens at one time

5. Be capable of passing through a standard 30-in. (0.75 m) door
6. Be reasonable in cost.
I1
I-i-:
35
VI,
so
--
x -AVE.
LIFE, FIRST FAILURE
0 -AVE.
LIFE, TURN TO TURN
~-AVE.
LIFE, PHASE
V1
'PHVARNS: OILMOIIEAHENICGLASS
'GOUD ORG
90
100
110
120
]I
-AN
IC-GLS
130
140 150
TEMPERATURE,
160
170 180 190
210
230
C 11/X SCALE)
Fig. 27-Life-temperature regression line of log average lives of motorette

system G tested at NRL
36
NRL REPORT 8095
100I
x -AVE.
LIFE, FIRST FAILURE
0 -AVE.
LIFE, TURN TO TURN
LIFE, PHASE
-AVE.
10
I-,
z
0
.5-
WIRE: POLYVINYL FORMAL

PHASE: ORG. VARN. GLASS
GROUND: ORG. VARN. MICA-GLASS

VARNISH: OIL MODIFIED PHENOLIC
.1I
90
100
110
120
130
140 150
TEMPERATURE,
OC
160
170 180 190
210
(1/K SCALE)
Fig. 28-Life-temperature regression line of log average lives of

motorette system G tested at NSRDL
37
230
-
Figure 29 is an artist's cutaway view of the chamber interior, illustrating the water
bath, cooling rack, specimen drawers, and mcunted specimens. The NRL pilot model
illustrated in Fig. 30 when assembled is 42 in. wide, 62 in. long, and 52 in. high (1.07 x
1.57 x 1.32 in). The chamber proper is 20 in. high and is mounted on a rack that
accommodates the heat exchanger, cuolant reserve tank, and circulating pump. The
double-walled cover is the only chamber outer surface that is insulated and is heated
slightly (and thermostatically controlled) to prevent condensation and resulting dripping
of water on the specimens.
Figure 31 shows the basic principle employed. The specimen rack is refrigerated by
a circulating coolant (water) that is thermostatically controlled to maintain a specified
temperature differential between the specimens and the surrounding chamber air. This
differential is independent of normal room ambient variations. Since both the heated
water bath and the coolant are thermostatically controlled, this independence is limited
only by the capacity of the system. Temperature control is not lost in the event that
the room ambient should rise to a temperature above that of the water bath. The heat
lost to the refrigerated rack keeps the water within the control of the heater, allowing
the balance of temperatures to be maintained. In case the room temperature falls below
that of the cooling rack, control is preserved by the heat st.pply of the water bath heater.
This balancing effect between heating and cooling systems eliminates the need for the
chamber to be in a temperature-controlled room, as a conventional dew-plus chamber
must be. The interior of the chamber was so designed that all motorette specimens would
be the same distance above the water bath and below the roof of the chamber, so that
specimens would be equally influenced by such factors as radiating surfaces, air temperature, and relative humidity. Figure 32 illustrates a specimen drawer and a set of 10
motorettes mounted on an open rack. The mounting rack greatly facilitates handling of
the specimens during the humidity, heat aging, and vibration cycles.
Figure 33 shows the rack of motorette specimens placed in the drawer with the
four.pronged quick-disconnect plugs in place. After the desired exposure to moisture, the
specimens are connected to a test stand by cables that lead to the receptacles on the
faces of the chamber drawers. A typical setup for this purpose is shown in Fig. 34. The
stand illustrated uses an improved NRL-designed test circuit that allows all components of
10 motorettes to be stressed simultaneously.
After the research and development on the condensation chamber was completed and
a commercial model made available, the Working Group embarked on a second round
robin motorette test program [111. Figure 35 compares the combined life-temperature
regression analysis curves for the 1958 original round robin and the 1968 reevaluation
round robin. It can be seen that the second round robin data yield a 15 0 C higher temperature index at 20 000 h than the original data.
In reporting on the second round robin, the Working Group stated "that it is
possible to obtain interlaboratory reproducibility with the proposed IEEE 117 Procedure."
In spite of the refinements and established controls, however, the test procedure in two of
the six laboratories varied enough that the results were considered outside the "test
family." Because of this the Working Group cautioned that "inter-laboratory tests must
insist on rigid adherence to test methods in all details if uniform results are to be
ach:eved."
38
{Ell
E-
NRL REPORT 8095
V,,
t-
)12
Fig. 29-Artist's cutaway view of condensation chamber
A further detailed study of the operating characteristics of the condensation chamber

was then conducted by the Navy. A time-dependent ratio of surface to bulk absorption,
which varied with the degree of thermal aging, was discovered. This phenomenon is
illustrated in Fig. 36, which compares the insulation resistance of a new motorette with
that of an aged one. Thus, caution should be exercised in adhering to the operating
condensation chamber operating specifications recommended in NRL Report 7469 [16].
As a result of this study, a minimum exposure time of 48 h was recommended, with
the difference in temperature between the motorettes and the air 25.5 mm (1 in.) above
maintained at 10 C.
A round robin test was then conducted between NRL and NSRDL to investigate the
degree of correlation that can be expected between two separate laboratories using the
condensation chamber as a moisture conditioning method. To eliminate variables other
than those contributed by the moisture conditioning cycle, both sets of motorettes were
fabricated and thermally aged at NSRDL. One set was transported to NRL after each
139
Fig. 30-Pilot model of NRL condensation chamber
aging cycle, to be exposed to vibration, humidity, and voltage stress cycles. The results
demonstrate that good correlation can be obtained between laboratories if the condensation chamber is used under rigidly controlled conditions. Details of the test conditions
and a breakdown of the data analysis are presented in Table 1. Plots of the regression
lines, log average lives, and 95% confidence limits are given in Figs. 37 and 38.
The main advantage of the condensation chamber with its controlled visible condensation is the approximately 2:1 savings in testing time for a given test temperature as
ccmpared to the Navy's method of no visible condensation. This advantage allows one to
test 100C to 150 C closer to the assigned classifying temperature of the insulation system
for the same test duration. For example, if when meeting the requirement that the
lowest test temperature be no more than 200C above the classifying temperature, the
Navy method produces an average life of 10 000 h. The condensation chamber method
would require only 5000 hours average life. In many circumstances, particularly in private industry, this 50% saving in testing time can be most important.
40
NRL REPORT 8095

COOLANT
TEMPERATURE
CONTROLLER
KIEATED COVER (INSULATED AND

THERMOSTATICALLY CONTROLLED)
TEST SPECIMENS
WATER BATH
TEMPERATURE
CONTROLLE
COOLING RACK
WATER BATH
HEAT EXCHANGER
RESERVE TANK
REFRIGERATIONHETRUNIT
Fig. 31-Block diagram illustrating basic principle of

condensation chamber
Fig. 32-Specimen drawer and set of 10 motorette specimens mounted on open rack
Since the Navy has already produced the great bulk of its motorette data over the
past 20 years using the no visible condensation condition, and equipment designed for
this method, it does not plan to make this major change at this time. However, the Navy
does plan to consider using the condensation chamber in the future as more background
experience and data are acquired on the newer, higher temperature insulation systems.
41
L
,--
Fig.33-ackof
inspeimend
mtorttespe
quic-diconnct
lac lugsin
42
. . ..........
NRL REPORT 8095
2I
Fig. 34-Test stand and cables used with condensation chamber for voltage
stress of motorettes
43
BRANCATO, JOHNSON, CAMPBELL AND WALKERI
'AW
-0
191 ROBIN
ROUND
Fig
100 110
12
190
210
230
44
NRL REPORT 8095

IO0.OTEMPERATURE SETTINGS'
MOTORETTES - 25* C
AIR (I" ABOVE)-260C
CHAMBER ROOF -300C
o
AGED MOTORETTE
Fig. 36-Effect of moisture conditioning on the resistance of the winding-toground insulation in an aged and an
unaged motorette
0.01
0001 0
12
16
20 24
28 32 36 40 44 48
52 56
TIME (HOURS)
Table 1-First Failure Data*

14000
1600C
1800C
ParameterL__________L__
Mean life (h)

Upper confidence limit
Lower confidence limit
Log average life
Arithmetic average life
Percent of standard deviation
Average standard deviation of
all Components
Average number of cycles
to failure
Types of first
failures and
number of each
type
_____
NRL
MELf
NRL
MEL
NRL
169
178
160
171
174
17.0
15.2
175
185
165
174
176
15.2
17.0
569
570
568
555
558
11.6
16.1
569
570
568
577
585
15.9
16.8
2169
2290
2045
2181
2150
9.5
10.9
2074
2194
195870
2050
2054
7.4
10.2
8.7
8.8
8.6
9.0
11.2
10.7
4
5
2
2
4
4
5
6
0
4
7
0
7
8
4
5* T-T Top
6 T-T Bottom
2 Phase
3 Ground
MEL
*The humidity cycle was 64 h; minimum drying time, 7 h. Heat aging cycles were 20, 65, and 192 h, with
0
corresponding aging temperatures or 180, 160, and 140 C.
IMarine Engineering Laboratory, now NSRDL.

iT-T means turn-to-turn.
45
JJ
__
0
U,
-\WIRE:
x-LOG AVERAGE LIFE, FIRST FAILURE
-.-
POLYVINYL FORMAL
PHASE: 10 MIL MYLAR/RAG
GROUND: 10 MIL MYLAR/RAG
VARNISH: OIL MUDIFIED PHENOLIC
46
11 E
TEMPRATRECC
.5
90
100
110
120
130
140 150160
TEMP[RATURE,
170
180
190
210
(1/K SCALE)
Pig. 37-Life-temperature regression line and 95% confidence limits of

motorette system RR2, tested in NRL condensation chamber
46
230
NRL REPORT 8095
10C~
50-
WIRE POLYVINYL FORMAL

10 MIL MYLAR/RAG
GROUND:1O MIL MYLAR/RAG
-PHASE:
VARNISH: OIL MODIFIED PHENOLIC
'A
x--LOG AVERAGE LIFE, FIRST FAILURE

0_
TE
5 6 1 0 10190
10 10 10 1354
90
ig
PRA
mooet
rersso
ytmR2
(1/
lin
SC-ALE)
an
5-ofdec
-8-~f-enpr
iiso
etdinNRLcnesto
hme
I47
m.
--
10
--
TREATMENT OF DATA
Calculation for Placing Regression Line
Kregression
A complete method for calculating the mean regression life-line, confidence limits,
line comparisons, and other statistical observations for thermal life data is
found in IEEE Standard 101-1972 [17]. However, for convenience the followng
method of calculating the regression line alone can be used when confidence limits and
other auxiliary information are not required. It is outlined in IEEE Standard 101A-1974,
which is an appendix to Ref. 17.
When the method of least squares is used, constant a and slope b of the regression
line may be derived by the following equations:
EZY- b Ex
N
=NXY2
N,2X
2X
-(X)
(1)
Y
(2)
where
X
reciprocal of the temperature, in kelvins
N = number of end-point values (hours of life) used in the calculation
Y = logarithm of the hours of life at a given temperature.

When this is solved for constant a and slope b of the regression line, temperature t
(in degrees Celsius) can be calculated by
t=
b Y-a
273.
(3)
For convenience, it is suggested that the log values for 20 000 h (4.3010 for logl 0
and 9.9035 for loge) and 1000 h (3.0000 for logl 0 and 6.9078 for loge) be used for Y in
Eq. (3). Either log1 0 or loge may be used, depending on the calculator used. These
values will provide enough space between the two temperature-point values for accurately
drawing in the regression line and are also conveniently located on the log hours scale.
Table 2 can be used in making the calculations. It gives the commonly used test
temperatures in degrees Celsius, their reciprocal values in kelvins, and the squares of these
reciprocals.
IA
48
NRL REPORT 8095
The sample calculation in Table 3 uses data for NRL twist combination 10, which
is found on Sheet 1, Column 3 (listing number) of Table Al. The hours-of-life data
represent the fifth and sixth failures at each test temperature, using the recommended
median truncated data method. However, the formulas can also be u,;ed for any number
of end-of-life measurements at any number of test temperatures. The oalculated regression
line and the individual hours-of-life data points at each test temperatu are given in Fig.
39.
Table 2-Temperatures and Equivalents

Temperature
(0C)
105
X2
Temperature
Xx
(X10 6 )
(C)10
0.002646
7.0013
185
0.002183
4.7655
125
0.002513
6.3152
190
0.002160
4.6656
130
0.002481
6.1554
200
0.002114
4.4690
140
0.002421
5.8612
220
0.002028
4.1128
150
0.002364
5.5885
230
0.001988
3.9521
155
0.002336
5.4569
240
0.001949
3.7986
160
0.002309
5.3315
250
0.001912
3.6557
165
0.002283
5.2121
260
0.001876
3.5194
170
0.002257
5.0940
280
0.001808
3.2689
175
0.002232
4.9818
300
0.001745
3.0450
180
0.002208
4.8708
320
0.001686
2.8426
6)
49A
UM,
XI
1-
10
TE
14
13
15
16
17
-87
PR TRxC(/KSAE
ad sxth
andfifhreresionlin
Fig.39-alclatd
falr1
ec
-a
21
19
ettmeatr
-50
ws
our-oflif
omiainn.1
230
NRL REPORT 8095
Table 3-Sample Calculation

Temperature
Life
140
5548
140
YX
0.002421
5.8612
8.6212
0.020872
5884
0.002421
5.8612
8.6800
0.021014
160
2410
0.002309
5.3315
7.7874
0.017981
160
2744
0.002309
5.3315
7.9172
0.018281
180
960
0.002208
4.8753
6.8669
0.015162
180
1046
0.002208
4.8753
6.9527
0.015352
0.013876
32.1360
46.8254
0.108662
X(XlO6
(h)
(0C)
E
N= 6
a Y - b 1X
= (6)(0.108662) - (0.013876)(46.8254)
(6)(0.000032136) - (0.013876)2
NZXY - ZXI Y
N2 X
(at 20b
(at 1000 h)
11.0533
46.8254 - (8154) (0.013876)
(2 X)
b
-
8154
9.9035 + 11.0533
8154
6.9078 + 11.0533
273
273
= 8154
116 0 C
181 0 C
Extrapolation Problems With Nonlinear Life Curves

The fact that problems exist when one attempts to extrapolate data from
accelerated thermal aging tests to temperatures in the operating range has been recognized by experts since the early days, when thermal aging tests were first considered as a
method for evaluating the thermal life of electrical insulation. More than 25 years ago
Dakin [6] pointed out that "if more than one chemical reaction proceeds simultaneously,
and if these reactions have different temperature coefficients, a plot of the logarithm of
the reaction rate constant or the time to reach a certain state of deterioration against the
reciprocal of the absolute temperature may not fall on a straight line."
51
i
By 1959 more than 10 years of accumulated data proved that the experts were correct in their early warning that all thermal life data could not be expected to fall on a
straight line. The Navy, for one, reported that in its study of motorette systems [101 it
found a break in the aging curve at 2000C for polyester magnet wire used in five different insulation systems tested over 3 years. This proof of nonlinearity is graphically
illustrated in Fig. 40. At about the same time, Saito and Hino [18] reported "that the
relation between log (life) and 1/T is not always a straight line." Their statement was
based on finding a break in the aging curve, all at the same temperature, for a given
film-coated magnet wire when comparing four different thermal evaluation methods.
(See Fig. 41.)
By this time several questions were squarely before us. For example: How near the
operating temperature need the lowest test temperature point be to reasonably assure us
that there is no break in the life curve? If a break appears in the curve at the test
temperature range, what should determine whether the data are to be disqualified for
extrapolation purposes? Can any other method of determining expected thermal life be
considered more reliable? These and other questions were the basis for confusion and
many misconceptions in the electrical insulation industry.
Even though extrapolation may lead to inaccuracies, the fact remains that it is the
best and most practical approach available today. It is clearly better than attempting to
make evaluations using general chemical classifications or using rule-of-thumb methods
like the 8. or 10-degree rules. It is obviously far more practical than waiting 10 years or
longer for field service trails before recommending or approving a new magnet wire for
a particular application. So, if extrapolation is desirable, it is essential to make it on the
most accurate basis possible while still taking a reasonable and practical approach to the
problem. With this goal in mind, procedures were outlined in the J-W 001177 Military
Specification [12]. Here the designated thermal stability test is ASTM D2307. The
lowest temperature test point must have an average life of not less than 5000 h and the
highest not less than 100 h. The spread between successive temperature points must be
at least 20 0 C, except where certain exceptions are allowed. Should the highest temperature point obtained yield less than 100 h average life, an additional point 10 0 C lower
may be obtained. Extrapolation to determine the classifying temperature (temperature
index) is based on the regression line of the three lowest temperature points.
Where nonlinearity exists the procedure graphically illustrated in Fig. 42 is followed.
The permissible departure from a straight line is gauged by the difference between the
extrapolation of the two lowest temperature points and the regression line of the three
lowest points. The differences are measured based on an arbitrary reference line of
40 000 h. (This is not a temperature classification line.) When the extrapolation of the
regression line of the three lowest points intercepts the reference line at a point that
exceeds both 20 000 h and 15 0 C over that obtained by extrapolating from the two lowest
temperature points, an additional temperature point is to be obtained. The time difference (measured vertically downward from the intercept point) and the temperature
difference (measured horizontally along the 40 000-h reference line) are represented by
distances A and B respectively in Fig. 42. This additional temperature point is to be
located at least 10 0 C below the lowest existing temperature point. To keep the testing
time as reasonable as possible, the procedure allows the additional point to be located
52
?A
NRL REPORT 8095
1,000
ASystem 10
X System 16
'V System 30
ol
System 12
O System 14
0
0
10
120
140
160
180
200
TEMPERATURE
220 240 260 280 300 320
*C (i/*K SCALE)
Fig. 40-Aging characteristics of polyester-type magnet wire

enamels used in five Navy motorette insulation systems
53
S
lO,O-
--
10,000
21,
z2
lt
120
140
160
180
I-
200 220 240 260 280 300 320
TEMPERATURE
*C (I/K SCALE)
Fig. 41--Comparison of four methods of thermal aging tests

using A-formal magnet wire 118]
10 0 C above the lowest existing temperature point if that point represents an average life
of more than 6000 h. However, if it is less than 6000 h the point must be located at
least 10 0 C below that point. After the additional temperature point is obtained, the
highest temperature point is discarded; the extrapolation is then based on the regression
line of the three lowest points.
This method for handling and interpreting the thermal-life data of electrical insulation appears to offer a good middle-course approach to the overall problem. The procedure is practical and simple and has been specifically designed to meet the needs of
electrical insulation evaluation. During the development of this graphical method of
testing for acceptable line&Aity a rather detailed application study was made using the
mathematical linearity test outlined in the IEEE 101 standard. In several cases the IEEE
101 method rejected data for nonlinearity while the Navy's graphical test passed the data
as acceptable. One consideration the Navy graphical method makes that the IEEE 101
54
NRL REPORT 8095
100
so-
AI
10-
0_
kAI
LINEARITY TEST
90
10C02104101010109
1e0
be
than120 000 h4
or0B mst 6be less than
0
55
230
method does not is where the lowest temperature point causes a break in linearity in the
in the
upward direction. From a practical engineering point of view an upward break
three
all
of
line
regression
the
line contributes to a conservative estimation of life when
Navy
the
where
43,
Fig.
in
seen
temperature points is extrapolated. This can clearly be
in
case
a
illustrates
44
Figure
test accepts the data and the IEEE 101 test rejects it.
another
is
45
Figure
data.
the
reject
test
101
which both the Navy and the IEEE
example of the Navy test accepting the data while the IEEE 101 test rejects it. In this
last case the line breaks downward but does not exceed the limit differences in extrapolation (150 C and 20 000 h at the 40 000-h reference line).
100
so
0o 10
MOTORETTE SYSTEM NO. 19

EMPLOYING POLYESTER WIRE
AND PHENOLIC VARNISH
IEEE 101 LINEARIT(' TEST: -0.13
NAVY LINEARITY TEST: A==-5000
90
100
110
120
130
I$0 160
140
TEMPERATURE,
170 180 190
210
230
C 11/K SCALE)
the Navy
Fig. 43-Example of identical motorette aging data accepted by
linearity test and rejected by the IEEE 101 linearity test
56
NRL REPORT 8095
100
--
So-m
-
COMBINATION NUMBER 3
EMPLOYING DBL DAGLAS SILICON
-T-IST
-
MAGNET WIRE AND NO VARNISHI

aA
IEE
10
iaiy0et
55
II
EARITYTEST: -2.5
IEEE-1-1-----
z-
TWIST COMBINATION NUMBER 47

EMPLOYING POLYVINYL FORMAL
MAGNET WIRE AND OIL MODIFIED
44
IEEE 101 LINEARITY TEST: -5.69

NAVY LINEARITY TEST: A= 17000 h
90
100
110
120
130
140
150
TEMPERATURE, CC
160
170 180 190
130C
210
tb/K SCALE)
Fig. 45-Example of identical twist specimen aging data accepted by

Navy linearity test and rejected by IEEE 101 linearity test
58
230
NRL REPORT 8095
Truncating Data to Shorten Thermal Aging Tests
k
:
The standard thermal aging tests for insulation materials (motorette tests and magnet
wire twist tests) call for aging 10 specimens until failure at each of three or four temperatures to obtain a life-vs-temperature curve. Valuable time can be saved if a valid estimate
of the life at each temperature can be made from data truncated at fewer than all 10 spec:mens. A statistical analysis was made, using 50 previous thermal evaluations, to study
the validity of three different truncated-data methods [19] : the log average of the times
for the fifth and sixth failures (method A), an estimate from a probability-plot fit to the
first five failure times (method B), and simply the fifth failure time (method C). The
methods wtere found valid, with method A the most and method B the least accurate.
All regression lines found by using method A fell within the 95% confidence limits of the
complete, or nontruncated, data method, and 90% of the results from the other two
methods fell within these limits. If method A had been used when the aging tests were
performed an average of 7.4 weeks, or 16.5% of the experimental time, could have been
saved with a loss in accuracy of less than 1%. Methods B and C would have saved an
average of 10.4 weeks, or 21.7% of the time, with a loss in accuracy of less than 2%.
It should be pointed out here that due to the intrinsic characteristics of the test proce.
dures, experimental error itself is conservatively estimated at between 5% and 10%. This
alone lends strength to the validity of the truncated data method and its use as a time-saving
mechanism.
As an extra precaution regarding the use of truncated data, a later investigation was
made of more aging testz, using data from some newer magnet wire insulations not included in the earlier study. Some of these wires yielded a wide spread of data at each
temperature point (up to 50% standard deviation) and nonlinear life curves with
unusually wide confidence limits. Others exhibited very linear life curves and reasonable
small spreads in data (10% to 20% standard deviation) resulting in extremely narrow confidence limits. In all cases truncated data method A gave regression lines that fell well
within the required limits of accuracy (0.20% to 1.78% error), and the aging time saved
ranged from 3 months to 1 year and 4 months. This additional study furnished conclusive
proof that the median (fifth and sixth failure) truncated data method is valid and can cope
with such extremes as highly nonlinear or widely dispersed data.
As a result of these studies it is recommended that aging tests using ASTM D2307
and IEEE 117 procedures be stopped after the sixth specimen has failed. This is
especially desirable for the lower test temperatures, at which it would take many months
or even years to obtain all 10 failures. In circumstances in which accurate confidence
limits are required for the regression line, the all-data regression analysis method is used.
However, a reasonably close approximation of the confidence limits can be calculated
by performing the confidence limit portion of the regression analysis on the basis of 10
failures at each test temperature. This is done by simply substituting the "N" value for
the number of specimens that were under test (N = 30 in place of N = 6 for a three
temperature point aging experiment). Also, a "student t" value, appropriate for the number of specimens under test and the percent confidence limit desired, is used. In the several cases in which confidence limits for all the data were compared with those for the
truncated data confidence limits, the percent of error did not exceed 3%.
59
AEROSPACE WIRES
Application of Thermal Evaluation Principles to
Aircraft Wires
"
Functional evaluation by the accelerated procedure was also used to determine

service-temperature ratios for insulated power cable and hookup wires used in military
aircraft. The program was begun at the request of the Air Force and was continued for
the Bureau of Naval Weapons Airborne Equipment Division. The results provided a
technology that will considerably reduce weight and bulk in today's aircraft and missiles.
To do their part in this effort, electrical engineers must use every available bit of information in designing equipment and circuits to meet these low
weight and bulk requirements. Yet in the interest of preservation of a very sizable investment, they must also
A
A
exercise
good
designing
to maintain reliability in these circuits throughout
the life of
eachjudgment
vehicle. inThus,
in designing
circuit wiring, particularly to carry temporary
overloads and pass through high-temperature zones, they must have more information
than is now available on deterioration rates of wire insulations, so that they can safely
cope with heating conditions above the normal operating temperature ratings of the
wires. For expedient application to these design problems, it was determined that such
information would be most useful as a graph of wire life vs temperature, accompanied
by a simple formula for summing up the deterioration of successive heating cycles to
determine the net life of the wire.
To place these objectives on a firm foundation, required a determination of whether
the philosophy of functional evaluation and the chemical deterioration rate equation
could be universally adapted to describe the characteristics of the many materials and
construction variations in insulations used in the aerospace industry for power transmission and hookup wires. By a comprehensive study of one class of wire (MI-W-5086)
in its many types of construction forms (using polyvinyl chloride as the primary insulation), the feasibility of this approach was confirmed. Extending the methods to other
wire classes using materials such as silicone rubber and polytetrafluoroethylene further
confirmed that the degradation rate principles could be applied, and it was demonstrated
that various constructions of a particular class of wire could be individually described.
Al
A
-A
Summary curves of each wire specification group that was studied in this program
are presented in Fig. 46. The 10 000-h intercept temperatures presented in Table 4 were
taken from these curves and were used to describe the classifying temperatures for each
wire insulation grouping described in the respective military specification.
60
~V
NRL REPORT 8095

100 000
50000
1m
__
Z>
L.
1687KC
F~!EE"
100
200
5000
8 2 6 0
TEMPEATUR
(RCPOCLASOUETMPRTRESAE
100 10
10101020
Tabl
(41
Tmeaue
4-n-cp
12vrgeC0siyn
j
MiliarySpecfictio
Tmpeatue at10 00A
MLW188,tympesaur
E and uve
EE
Iwir61
ir
290iiato
-.
41,
Cyclic Temperature-vs-Life Calculations
Most wires in aircraft will, in actual service, operate at a variety of temperatures

caused by intermittent current loads and changing ambient conditions. The curves of
Fig. 46 can be used to predict the effects of this temperature cycling on insulation life.
To compute the life for any heat aging cycle, we find it convenient to treat the relative
aging in the manner suggested by Sumner, Stein, and Lockie [201. This method assigns a
factor of unity to the operations at any temperature other than the reference temperature. The relative rate of deterioration is inversely proportional to the computed thermal
life at that temperature.
Dividing the logarithmic expression of the chemical rate equation at reference
temperature T o by that at operating temperature T1, one obtains the equation of the
aging factor-vs-temperature curve as expressed by Whitman [211:
log R
logB
1)
or
-5
R
where
R = relative aging factor
B = constant of the material
L 0 = life at reference temperature T o
L, = life at operating temperature T 1
To , T,
- temperature in kelvins (0C + 273).
Consequently, a relative aging factor-vs-temperature curve derived from a straight-line

life-temperature curve is also a straight line when plotted on the same coordinate paper,
but has a slope of opposite sign. Thus, when the regression curve of life vs temperature
for a wire is available, aging time at one temperature is to converted to the equivalent
aging time at another temperature by applying the relative aging factor principle.
Suppose we wish to use a wire having the life-temperature curve illustrated in Fig.
47. The life at 1050C is 10 000 h. Figure 48 is the curve of the relative aging factor
vs temperature, derived from Fig. 47. The factor is 1.0 at 1050C, the rated maximum
continuous temperature of this fictitious wire.
62
" " +"PF-+

'+" "
' "--'"
= m'++'--"?+:'
" '+"
..-.-
+: .....-T +- _.+r + -:
. ..
"-'-r "
+I
NRL REPORT 8095

100000
b0000 k
LIFE.vs.TEMPERATURE CURVE
(OF A FICTITICUS CABLE)
1000
10100-- _
curve of
Fig. 47-Life-vs-temperature
a fictitious wire insulation"_
...._-
RATEA+SIT
RULE
CHEMICAL
LIFE-LOG
LOG
L
LO
_--__
100
100
120
140
240
160 180 200
TEMPERATURE ('C)
280
320 360 400
ABSOILUTE -TEMPERATURE SCALE)

(RECIPROCAL
1007
AGING FACTOR vs TEMPERATURE
LRELATIVE
'OF A FICTITIOUS CABLE

INSULATION)
@ 105C
=f1.0
SR
insulation
~U
Rt =LIFE O 105
LIFE 0 T 1
cc1.0
?
'_100
0.10
-.
120
140
10
18D 200 220 240
280
320
30D 400--
TEMPERATURE 1 C)
(RECIPAOCAL ABSOLUTE-TEMPERATURE SCALE)
63
Fig.temperature
48-Curve offorrelative
aging factor
wire
a fictitious
vs
Applications
Once the regression curve of life vs temperature is established and the curve of the
relative aging factor vs temperature is derived, it is possible to predict the effects of
temperature cycling on the life of the insulation. An example of the application of this
principle to reduce wire weight would be the selection of the maximum wire size for an
aircraft electrical circuit required to carry cycles of temporary overloads, such as during
landing gear operation. Suppose the specifications require that the reduction in life of
the wire due to the overload must not be less than 20% of the life at the rated temperature, and the following performance conditions are specified:
1. Normal current at rated temperature of 105 0 C (MIL-W-5086A, type I wire)
2. Allowable current during overload state
3. Portion of operating time at overload state = 5%.
To find the maximum steady-state temperature to which the wire insulation can be heated
during the overload cycle, let
to = time of operation at 105 0 C
tI = time of overload state = 5% of total time
A-1
T, = temperature at overload state

R, = relative aging factor at T1
100 = percent life at rated temperature of 105 0 C.
A-
Expressing the total life available as

to + Rltj
100%,
and the specified minimum life as

to + ti
80%,
solve for R 1 by substitution:

(80
tj) + Riti
20+1
64
5.
5.
100
NRL REPORT 8095
If we assume that the curve of relative aging factor vs temperature in Fig. 29 represents
this type of wire, temperature T1 at R, = 5 is 1280C. The maximum wire size that does
not exceed a temperature of 128 C while carrying the specified overload current can then
be selected (from the time-temperature-current curves representing this type of wire) and
installed.
The overload cycle in this example has been treated as rectangular in waveform,
assuming that the time of temperature rise is relatively short in comparison to the total
time at the maximum temperature during the overload. The problem of a transient
temperature rise of a relatively short duration, such as might occur due to a cir ,,;t fault,
can be solved similarly by expressing the temperature rise and fall vs time as e). ,, ential
curves. The relative aging time of the cycle would then be calculated by integrating the
curves of the instantaneous relative aging vs time.
Three different experiments were conducted in this program with the heat exposure
cycles at different schedules and temperatures, and in each test the R-value as calculated
in the above example was experimentally verified. Thus, it was demonstrated that it is
possible to apply the relative aging factor calculation to convert aging temperature cycles
to equivalent aging at any one temperature and thus predict the life of wire for cycling
service conditions. When overloads, causing rises to higher temperatures, occur in known
service cycles, it is possible to calculate the number of cycles that can be safely expected
of the wire.
FUTURE WORK ON PROBLEMS IN NEED OF SOLUTION
Today thermal aging technology is approaching maturity and is being used by both
industry and government in insulation systems design and materials specifications. The
universality of application, however, means not that the problems have all been
eliminated, but merely that the need is so great that an imperfect tool is better than
none.
In the main, what has been accomplished to date is the develop:nuit of test procedures that define and simulate a functional model of the insulation s)sem of interest,
and correspondingly of the appropriate environment and failure criterion. These procedures have permitted the development of standards for obtaining index ratings of
materials and classifications ratings for systems. This obviously aids and encourages
design of new systems.
While progress in thermal aging has been significant, the problems remaining are
equally significant. The authors commend the following four areas for continued exploration.
65
kHumidification
The humid environment is still one of the most troublesome to simulate and equally
difficult to standardize. Industry has used humidification with condensation with
resulting large scatter in results. The Navy, in the effort to reduce the spread in data, has
used humidification without visible condensation. While the latter has yielded reproducible results, it has been at the cost o' ,a-reased testing time. This impasse was
, '.the development
resolved in the IEEE 117 Working G.o ?, .tY
of the condensation
this
pr
tides considerably improved humidification,
Whiie
chamber previously described.
problems remain. Some of these problems are che ratio of surface humidification to bulk
humidification, the duration of humidification, and the method of measuring. When
these questions are answered, some standardization of humidification can take place.
Failure Criteria
Failure criteria may have to be broadened for specific use; aging phenomenon may
not always dictate the end of life of insulation systems. Voltage breakdown due to surface tracking or corona, mechanical rupture or deformation of insulation, and deterioration due to an inimical chemical or to dust environments may well determine the end of
useful life before full thermal degradation sets in.
Duration of Tests
Testing at present takes entirely too long and is too expensive. Some aging results
are influenced by the overly high temperature of tests, at which chemical kinetics are not
the same as at operating temperatures. Work should now turn to measuring the degradation near or at operating temperature for a relatively brief period (hours or a few days)
and integrating the rate over the expected life of the system.
Combined Environments
Ichains
In addition to the thermal environment, insulation life is affected by such other

environments as chemical fumes, vacuum, and radiation. The advent of the nuclear and
space age has injected the parameter of radiation into design consideration. It is important to know and be able to measure the effects of radiation on the electrical and life
properties of insulation. These effects are not always predictable because radiation
induces simultaneously two opposite events in a polymer; on one hand it links polymer
together, increasing the molecular weight by "cross linking," while on the other
hand the polymer chain is fractured by "chainscission," causing degradation. The rate
balance between these two events determines if a material improves or degrades in some
specified characteristic. However, this equilibrium point is also a function of temperature.
Thus, while a material at a given temperature may be improved in some of its characteristics by radiation, the same material under the same radiation conditions may degrade if
the temperature is changed.
66
NRL REPORT 8095
This was demonstrated by Campbell [22] when he exposed magnet wire to gamma
radiation, then aged the wire at a given temperature. He repeated these exposures on
similar wires but combined the environments; the results were very different. For
example, the simultaneous aging at high temperatures in gamma radiation of a polyimide,
a polyvinyl formal, and a polysiloxane produced considerably longer lifetimes than
simple thermal aging at the same temperature. On the other hand, polytetrafluoroethylene
deteriorated much more rapidly under simultaneous heat and radiation.
This means that test conditions must be designed to simulate service environments,
if results are to be meaningful. The synergism of heat and radiation cannot be overlooked. While radiation was used as the second environment in illustrating this point, this
applies to any combination of environments. Effort should be devoted to expanding test
procedures toward combined environments where physical, chemical, mechanical,
humidity, vacuum, and radiation conditions, etc., are singly or in combination the prime
cause of failure.
In conclusion,
one can
infer that the posed questions in themselves reveal the
tangible
and considerable
progress
that has been made in this field. The statement of
remaining problems is not so fundamental as to suggest that little has been done; they are
on the other hand specific enough to reveal the considerable command the profession
has in thermal aging. It is hoped that satisfaction with the present state of the art will
not delay or set aside investigation of these and other related issues.
CONCLUSIONS AND RECOMMENDATIONS
As a result of this study the following conclusions and recommendations are made.
1. The magnet wire twist and motorette insulation systems procedures have had
sufficient successful history, in both industrial and governmental laboratories, to be
continued as standard procedures.
2. The humidity cycle is a reliable diagnostic tool in determining the end of thermal
life of motorette insulation systems. However, the following considerations should be
taken into account.
a. Specimens should be exposed for at least 60 h when humidity without
visible condensation is employed. However, the minimum exposure time can be reduced
to 48 h when condensation chamber is used.
b. Years of Navy experience have proved that humidification without visible
condensation provides a reliable method for obtaining uniform and reproducible results in
evaluating a wide range of insulation systems.
c. Where overall aging time is to be minimized, humidification with visible
condensation is recommended provided the IEEE 117 Test Procedure is strictly adhered to.
3. The median method for truncating data has been proven valid; it saves substantial
aging time (up to 20%). It is strongly recommended that this time saving method be used in
future thermal aging studies.
67
4. As a result of recent field and laboratory experience the Navy's classifying lifeline benchmark should be established at 20 000 h, in lieu of the original 40 000 h, when
considering data from the magnet wire twist test and the motorette procedure using
humidification without visible condensation.
5. In future work the following should be pursued.
a. Most electrical insulation is, in reality, subject to combinations of aging
factors, such as voltage and temperature or radiation and temperature. It is strongly
urged that an adequate understanding of the pnysics and chemistry of aging be acquired,
so that multistress aging can be predicted and evaluation procedures developed.
b. The interrelationships or mutual influences of materials in the presence of
others should be studied. Experience indicates that properties of combined electrical
insulations do not necessarily reflect the characteristics of each material separately.
c. It is recommended that more sensitive measuring techniques be explored.
This should be aimed at assessing the rate of aging close to the operating temperature
and to shortening test periods (possibly to one month).
d. Further study of humidity should be pursued, particularly in the areas of
surface-to-bulk ratio effects, simplifying procedures and increasing reproducibility.
e. With a better knowledge of aging phenomena, possible nondestructive evaluation procedures should be explored and developed.
ACKNOWLEDGMENTS
The authors express their sincere thanks to John F. Tobin, Naval Ship Engineering
Center; Robert J. Flaherty and Jack A. Kimball, Naval Ship Research and Development
Laboratory; and Robert S. Phillips, Curtiss L. Bagget, Maxwell E. Sharp, and Robert J.
Lambert, Naval Research Laboratory. These people, through their many contributions
and supporting efforts to the insulation program, have made the successful achievements
to date possible.
REFERENCES
1. C. P. Steinmetz and B. G. Lamme, "Temperature and Electrical Insulation," AIEE
Trans. 32 79-89 (1913).
2. V. M. Montsinger, "Loading Transformers by Temperature," AIEE Trans. 49,
776-792 (1930).
3. A. T. McClinton, E. L. Brancato, and R. S. Phillips, "Effect of Voltage Stress and
Vibration on Insulation Life," NRL Memotandum Report 207, Sept. 18, 1953.
4. C. G. Currin and J. F. Dexter, "A Method for Evaluation of Thermal Stability for
Magnet Wire Enamel," AIEE S-61, p. 26-30, June 1954.
5.
G. A. Cypher and R. Harrington, AIEE Trans. 71, Part III 251 (1952).
68
A4
NRL REPORT 8095
6. T.
W. Dakin,AIEE
"Electrical
Treated as a Chemical Rate Phenomenon,"
Trans., Insulation
Part 1, 67,Deterioration
113-122 (1948).
7.
"Tpst Procedure for the Evaluation of the Thermal Stability of Enameled Wire on
Air," IEEE Standard 57, Jan. 1959. Superseded by American Society for 'resting
and Materials, "Test for Relative Thermal Endurance of Film-Insulated Round
Magnet Wire," ASTM Standard D 2307-68.
8. International Electrotechnical Commission "Recommendations for the Classification

of Materials for the Insulation of Electrical Machinery and Apparatus in Relation to
their Thermal Stability in Service," IEC Publication 85, first edition 1957.
9. S.J. Antalis and G. M. Bell, "Predicting Transformer Life Through System Technology," Insulation/Circuits,55-60 (Apr. 1972).
10. E. L. Brancato, L. M. Johnson, and H. P. Walker, "Functional Evaluation of
Motorette Insulation Systems," Elec. Manufacturing, p. 146.153 (Mar. 1959).
11.
L. P. Mahon, "A Re.evaluation of the Life Expectancy of Class A Random-Wound

Motor Insulation as Determined by the Proposed IEEE 117 Test Procedure," IEEE
Trans. 68, TP 636 PWR.
12. Interim Federal Specification, Wire Magnet, Electrical, J-W-001177 (NAVY.SHIPS),

Feb. 13, 1974.
13.
R. J. Flaherty and H. P. Walker, "Thermal Life of Enameled Round Magnet Wire

with Various Twist Holder Configurations," IEEE 32C79-63, in Proc. Seventh Elec.
Insulation Conf., Oct. 15-19, 1967; published by IEEE, New York, 1967.
14.
"IEEE Standard Test Procedure for Evaluation of Systems of Insulating Materials

for Random-Wound AC Electric Machinery," IEEE Document 117, 1974.
15.
L. M. Johnson, "An Analysis of the Intrinsic Characteristics of the AIEE Round

Robin Motorette," AIEE CP 61-349, Feb. 3, 1961.
16.
L. M. Johnson, "Humidification in the Thermal Life Determination of Motorette

Insulation Systems," NRL Report 7469, Sept. 7, 1972.
17.
"IEEE Guide for the Statistical Analysis of Thermal Life Test Data," IEEE Docu.
ment 101-1972, Mar. 17, 1972.
18. Y. Saito and T. Hino, "Study of Thermal Deterioration of Enamel Wires by Mass
Spectrometer Method," AIEE CP 60-88, AIEE Winter Meeting, 1960.
19.
L. M. Johnson, F. J. Campbell, and E. L.. Brancato, "Statistical Analysis of

Truncated-Data Methods to Shorten Thermal-Aging Tests of Electrical Insulation,"
"NRL Report 7468, Sept. 19, 1972.
20. W. A. Sumner, G. M. Stein, and A. M. Lockie, "Life Expectancy of Oil-Immersed

Insulation Structures," AIEE Trans. 72 (Part II), 924-930 (1953).
21.
L. C. Whitman, "Loading of Sealed Dry-Type Transformers," AIEE Trans. 76 (Part

III), 214-223 (1957).
22.
F. J. Campbell, "Combined Environments versus Consecutive Exposures for Insulation Life Studies," IEEE Trans. NS-11 (5), 123-129 (Nov. 1964).
69
-_2
i
Appendix A
USE OF COMPUTER READOUT TABLES Al AND A2
Table Al provides information on the computer regression analysis of temperaturelife data for magnet wire twist evaluations, and Table A2 provides the same information
for motorette systems. As an aid to finding specific magnet wire combinations or
motorette systems in the two tables, indexes are furnished. The first column of the
index lists the sheet (page) number of the table and the second column the listing
number for each twist combination or motorette system. The fourth and fifth columns
list the index key numbers to the generic names of the magnet wire insulations and
varnishes used in each twist combination and motorette system. These are given in
Tables A3 and A4. Each sheet of Tables Al and A2 has 12 listing numbers, 6 on the first
line of the top half and 6 on the first line of the bottom half. In columns under these
numbers are given the three lowest test temperatures (1, 2, and 3), the 95% lower-only
confidence limit at these three temperatures, and the mean regression line hours for
temperature 3 (furnished to aid in plotting the regression line). Below this are the
regression line temperature index values for 20 000, 30 000, and 40 000 h, along with
the corresponding values for the lower-only confidence limit.
While 254 twist combinations and 62 motorette systems are listed in Tables Al and
A2, the number evaluated was greater than this. Many of the earlier evaluations did not
meet the more stringent requirements later set forth by the Navy for thermal evaluation
of electrical insulations, particularly in terms of the minimum hours of life at the lowest
test temperature and the additional temperature points required for extrapolation when
the aging data did not meet the linearity requirements. In addition, some of the twist
combinations and motorette systems were used in special side studies, and this data did
not lend itself to the type of processing suitable for the tables.
It will be noted that in Tables Al and A2 the 95% lower-only confidence limits
were calculated and listed in favor of the 95% upper and lower confidence limits. This
was done, as recommended in IEEE 101, because it is more meaningful to determine to
a 95% degree of confidence the lower bound of the mean regression life at a given
temperature. Since the lower-only confidence limit yields a longer life at a given temperature (or a higher temperature for a given life) for the same degree of confidence than a
two-sided confidence limit it is more favorable to use the lower-only limit.
70
&g
NRL REPORT 8095
Index to Computer Data Sheets (Table Al) and Twist Sample Identification
Sheet
Number
Listing
Number
1
1
1
1
1
2
3
4
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
Combination
Number
Wire
Insulation
V
Varnish
AWG
Size
8
9
10
46
A
A
A
A
NONE
2A
2B
2D
20
20
20
20
51
52
11
47
135F
135B
135C
09
16
17
18
45
65
66
67
123
124
125
126
127
71
72
73
68
69
70
98
99
82
100
161
83
272
273
274
275
276
277
278
198
A
A
A
A
A
A
A
B-I
B-1
B-1
B-1
B-1
C-1
C-1
C-1
C-1
C-1
C-1
C-I
D-1
B-4
B-4
B-4
D-2
D-2
D-2
D-2
D-2
D-2
D-2
D-2
D-2
G-1
G-1
G-1
G-1
G-1
G-1
G-1
D-2
2F
2F
5A
2D
2A
2A
2A
2A
NONE
3A
3A
4A
NONE
4A
10A
NONE
4A
NONE
4B
NONE
NONE
4A
10A
NONE
4A
10A
NONE
10A
4E
4E
4E
(2 DIPS)
NONE
4E
4D
5E
4C
2C
4F
5C
18
18
20
20
18
18
18
18
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
18
18
18
20
20
18
20
20
18
20
20
20
20
20
20
20
20
50
71
2F
18
FIndex
Et
RE
ME
to Computer Data Sheets (Table Al) and Twist Sample Identification (Continued)
Sheet
Number
Listing
Number
4
4
4
5
5
5
5
5
43
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
5
5
5
5
5
5
6
6
6
6
6
6
6
6
6
6
6
6
7
7
7
7
7
7
7
7
7
7
7
7
8
8
8
8
8
8
8
Combination
Number
119
321
322
373
340
341
342
343
344
303
305
306
307
308
309
310
361
362
259
260
261
262
263
264
265
266
283
285
286
267
268
269
270
287
288
253
252
251
250
249
356
357
358
359
232
233
72
IN
Wire
Insulation
Varnish
AWG
Size
D-2
G-1
G-1
G-1
G-1
G-2
G-2
G-2
G-2
H-1
H-1
H-1
H-1
H-1
H-1
H-1
H-1
H-1
G-4
G-4
G-4
G-4
G-5
G-5
G-5
G-5
G-5
G-5
G-5
G-6
G-6
G-6
G-6
G-6
G-6
P-1
P-1
P-1
P-1
P-1
H-6
H-7
H-8
H-9
P-2
P-2
5D
NONE
NONE
9A
NONE
4E
2E
6B
10A
11A
NONE
NONE
NONE
NONE
4E
4E
4F
9A
NONE
4E
10A
7A
NONE
4E
10A
7A
6B
6B
6B
NONE
4E
10A
7A
6B
2D
4E
6B
10A
2D
NONE
NONE
NONE
NONE
NONE
NONE
2D
20
20
20
19
18
18
18
18
18
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
18
18
18
20
20
20
20
20
20
NRL REPORT 8095
Index to Computer Data Sheets (Table Al) and Twist Sample Identification (Continued)
Sheet
0-Number
Listing
Combination
Wire
Vrih
AWG
Number
Number
nsulation
Vrih
Size
10223
1A2
10
10
10
10
10
10
113
114
115
116
117
118
885
294
295
32
281
22
P-2
H-2
P-2
P-2
B-2
B-2
63B
4EA
7A
7A
5A
6BN
20
36
20
20
20
20
10
119
289
B-3
NONE
20
9i
91
9i
9i
11
9i
9i
9i
10
10
10
12
12
12
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
298
299
345
346
347
348
39
350
351
32
353
354
355
254
279
B-3
B-3
K-i
K-1
-1
K-i
H-1
H-1
H-i
H-i
H-i
H-2
H-1
E-2
NOE
NON
NONE
40A
2DB
1B
13A
NOE
NOE
4EA
13C
NONE
1B
NONE
4E
24
24
18
18
18
18
18
20A3
20
20
20
20
20
18
18
12
136
329
E-2
7OA
18
12
298J-1
10
11
NNE73
W-
Sheet
Listing
Combination
Wire
Number
Number
Number
Insulation
12
12
12
12
12
12
12
12
13
13
13
13
13
13
13
13
13
13
13
13
14
14
14
14
14
14
14
14
14
14
14
14
15
15
15
15
15
15
15
15
15
15
15
15
16
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
330
371
372
374
375
363
364
365
217
237
238
239
240
241
242
338
290
291
185
186
188
189
313
314
315
317
300
254
255
256
257
258
132
133
148
177
293
139
140
141
154
155
156
157
158
M-1
M-1
N-1
M-1
M.1
H-4
H-4
H-4
H-5
0
0
0
0
0
0
0
H-10
H-10
I
I
I
I
1
I
I
I
M-2
M-2
M-1
M-2
M-2
M-2
H-1
H-1
H-1
H-1
H-1
E-1
E-1
E-1
F-2
F-2
H-1
H-1
H-1
Varnish
AW
Size
7A
10A
10A
NONE
NONE
NONE
4E
10A
NONE
NONE
6B
7A
4E
2D
10A
12A
NONE
NONE
NONE
10A
4E
14A
NONE
NONE
NONE
4E
7A
NONE
2D
10A
6B
4E
NONE
13B
8A
13C
10A
NONE
2C
10A
NONE
10A
NONE
9A
10A
18
20
20
20
20
18
18
18
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
18
20
20
32
23
23
23
20
20
20
20
20
74
Fri-
NRL REPORT 8095
[.
Sheet
Number
Listing
Number
Combination
Number
Wire
Insulation
16
16
16
16
16
16
16
16
16
16
16
17
17
17
17
17
17
17
17
17
17
17
17
18
18
18
18
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
159
160
292
95
96
97
108
109
57
84
85
92
366
368
370
367
55
56
101
102
103
104
105
106
107
116
117
H-1
H-1
H-1
D-3
D-3
D-3
D-3
D-3
B-7
H-3
H-3
H-3
L-3
L-4
L-5
L-2
B-7
B-7
S
18
209
18
18
18
18
18
18
18
19
19
19
19
19
19
19
19
19
19
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
AWG
Size
S
T
T
U
U
U
S
6B
4E
10A
NONE
10A
4A
NONE
10A
4D
NONE
10A
13A
NONE
NONE
NONE
NONE
4D
4D
NONE
4A
10A
NONE
10A
NONE
10A
NONE
4A
20
20
28
20
20
20
20
20
18
20
20
20
20
20
20
20
18
18
20
20
20
20
20
20
20
20
20
118
10A
20
119
120
121
122
130
131
136
137
138
2
3
34
07
013
014
12
13
T
T
U
U
Q
Q
B-6
B-6
B-6
W
W
W
V
V
Y-1
Y-1
--1
NONE
10A
NONE
10A
NONE
2C
NONE
2C
6C
10A
4D
NONE
10A
2A
10A
NONE
6A
20
20
20
20
20
20
20
20
20
18
18
18
20
20
20
20
20
75
-~
.----
------
Sheet
Number
Listing
Number
Combination
19
19
20
Number
Wire
Insulation
_____j
AWG
Sz
227
228
229
14
15
010
Y-2
Y-2
Z.1
NONE
10A
5A
20
20
20
20
20
20
20
20
230
231
232
233
234
22
58
59
60
61
Z.1
Z-2
Z-2
Z-2
Z-2
6A
NONE
2A
20
20
18
18
18Al
20
20
235
236
62
80
Z-2
Z.3
6A
10A
NONE
18
18
18
20
20
20
20
21
21
21
237
238
239
240
241
242
243
81
017
018
022
023
024
23
X
X
X
C-5
C-5
C-5
2G
10A
2A
2A
5A
10A
NONE
18
18
18
20
20
20
20
21
21
21
21
21
-21
21
21
21
22
22
244
245
246
247
248
249
250
251
252
253
254
31
32
41
43
44
74
75
78
79
128
129
B-5
B-5
B-5
C-2
C-2
R
B.
Y-2
Y-.2
C-4
0-4
NONE
2B
2A
4E
NONE
NONE
15A
10B
NONE
NONE
20
20
20
20
20
20
20
20
20
20
20
20
Z-3
Vih
-4
IA
76
--
NRL REPORT 8095
Table Al-Computer Readout Data, (Sheet 1)

LITING ,UMbE
GCUim1'JAFIO,'N NultjEli
10
46
50
51
3 LOJWEb
FKST
rE14?1EHATUHES
1
2
3
120
140
160
130
140
160
140
160
180
140
160
180
140
1 60
160
140
160
180
LJG. AVG. LIFE

Arrt . F
TEMP. (H3.
1
2
3
7644
2000
582
I1584
46?0
1402
5714
2572
102
4788
158"1
756
3864
756
612
4216
1558
566
9t/.
L.C.L.UJLY
VAL UE.(HS. )
SFE6
. l,2&3
I
2
3
6903
191d
t27
9412
4643
1268
5532
2332
975
42U9
1691
663
2318
1066
363
416
1492
565
XEU. LINJE
564
1357
I042
720
485
579
H0Ut
TEMP I,'DEA
VAL.( C)
L-C.L-
107
104
1P0
119
116
113
113
109
107
88
114
113
3U,000 HLJif EIMt'

.INU EA
F&'1?*.
VAL.( C)
L.C.L.
102
99
115
113
109
1U
106
10 2
I00
79
107
106
40O00 HOUR( rEMP . INL)EA

IEMIJ. VAL.( C)
L.C.L.
98
95
III
1I10
103
I00
1I01
97
95
74
103
102
MEvAJ
20,0POO
IEM'".
LI.F1~~ N~18~
Mdi,)mlAI
47
,' NUME'
3LOWESr
3
E~AF K
F F~W
79
AVG.
0
LEA
T EMeIN
20,000 Ho0t
L.C.L.
C)
VAL-(
TVE~1t.
I ,
L. C)
HVM
0
EAr
L)~)~
Il N00U
i
L.C.
iriJEA
rt-r
4J,0O0
L. .L
MJ.( C)
4J UO
I3tC
14U
160
11301
140
160
0
6039
1452
4
4692
124
612
t)365
1365
503
333U0
13 3
60
tk
05~
3596
10 5 1
794
3U
1 1 37
434
16 0
54
39U5
1442
507
4726
1446
44 2
12 6b
619
4130
1216
565
600
Iu
104
133
1263
125
122
120
115
115
110
120
1113
97
127
21
116
114
114
1 13
101od
103
I15t
112
96
9 ?
121
11-)
114
109
304
104
99
III
106
LIFE
A EI21564 3f
fESd
~iAJrE.LINE
3
135b
62
3696
j13
13400
13F
11219
5162
109
16U
9.L.L.J'JL'Y
)
V LL (;.L
V A U Z ( K &) ),
4 140
16U
160
11
160
IdO
7
LOJG.
Ii
0U
160
Id u~
14
10
77
DO
if
MARK ON 'HIS 4H F
0',j' O~
VD"LAY '5"
BRANCATO, JOHNSON 1 CAW]L
D1WALKER
Table Al Computeiraeadlout Data, (Sheet 2) (Continued)

13
14
15
16
17
Is
09
16
17
is
45
65
I
2
3
I16U
180
200
1S0
200
220
16U
tau
200
160
180
200
t8u
200
221)
2CR0
220
24U
LOG* AVG*. LIFE

AT TEST
TEMP*(HiRS*)
1
2
3
15735
4894
2742
9912
3716
552
28572
7056
3288
29740
7362
3116
10344
36849
912
13752
2684
227'
95.4 L#CtL#O*LY
VALUESC MRS*)
OThiIPS- 1#2&3
1
2
3
12818
5564
2224
925d
2419
526
23194
6100
2656
25211
8246
2631
9896
3059
882
12752
1774
215
SK~AN KEG* LINE
~3
2509
662
2914
2891
1004
270
20#000 HOUR TEMI'.INOEX

TV-Me. VAL*( W) L6eC.L&
30#000 HOiURt
FEMP*~INDEA
TE*VA6#( W) L*C.Lo
153
ISO
145
141
173
169
166
163
165
163
1Sb
155
165
164
159
157
171
168
166
162
198
196
194
192
40.000 HO0UR ;CPOINVEA

TE3e. VAL*.( CW le#C*L.
140
134
165
159
153
ISO
154
152
162
156
192
189
LISTIN4G NUMB~iER
L.COMtBINATION
AtUaER
3 LOWESf
TES'C
4TEMPFAATURES
19
20
21
'2
23
24
66
67
123
1~
125
126
L1i. AVG* LIFE

'AT IiST
1
2
3
1
2
200
220
240
10576
3970
190
200
220
14200
6534
200
220
240
12270
1425
20u
220
240
8736
2473
200
220'
240
155571
1848
200
220
2-40
7981
1596
t 95% LpC.L.'fNLy
VALUtSCRS.)
ThS.1*il&3
1
2
3
9746
2347
442
11516
7386
2669
10133
1609
256
8419.
2062
494
11409
2275
411
7483
1639
382
MEAN REG, LINE
585
3011
286
548
490
399
204000 HOUR TEP*INDEX

TMRP
VAL-( C) L.C-L.
195
140
181
17d
194
193
190
188
196
193
189
188
TEMP 6I1NOEK
190
174
190
185
191
184
184
169
189
183
169
16
188
186
182
179
188
Ids
185
183A
lt7
&81
LISTING NUMBER
CdMBINAT14.4 NUMBER
LOWEST
EST
AN
__TEMEATURES
3O000OO
TEMP
4UR
VAL*( C)
L*C*L.
40i,000 HOUR fEMP-INDEXI

TEO&P VA"..( C) L.C*L.
Ek=-.
If
181
180
--
7
~
P-
'T-
IN
R"I
?ML R!R'RT 8095

Table Al-Computer Readout Data, (Sheet 3) (Continued)
LISTING NUMBER
COMBINATION NUMBER
25
26
27
26
29
30
127
71
72
73
68
69
3 LOWEST
TEST
TEMeERATURES
1
2
3
200
220
240
180
200
220
1140
200
220
200
220
240
190
200
220
180
200
220
LO8GO AVG* LIFE

AT TEST
TEL4POCHRSO)
1
2
3
9509
1848
408
45882
3925
2572
27772
4357
2959
7403
2047
102
13933
4493
1342
20606
2744
1352
951 L&C.L.84L-Y
VAUECIRS)
*TEMPS. 1*2&3
1
3
9488
146
407
2i008
648
1213
16374
6191
1724
6532
943
93
10711
5221
1071
13014
3726
845
4MN EU* LINE
408
1844
2332
141
1249
1099
20#000 HOUR~ TEMI'.INDEA

TEMP. VAL*( C) L#0#L&
191
191
187
182
182
176
194
155
184
Idl
178
173
30.000 HOUR TEMPO 1 DEA

TEMIPO VAL*( C) L.C.L.
187
187
182
175
175
167
191
179
173
1-4
176
167
184
I14
178
171
188
176
170
169
161
182
172
162
40#000 HO0UR
Tf.DIP
VAL#(
C)
TEM4P. INDE
Li.C.L
31
32
33
34
35
36
70
YO
99
62
100
161
1
?"T
3
2UU
22U
240
2uU
220
240
2U
220
24U
200
220
240
200
220
24U
200
220
240
1
3
6959
1492
dU3
9036
2924
'04
970d
2356
360
6'u2
1564
247
7632
2356
624
6644
22'aI
576
95. L.C.L.UALY
VALUES(HtrS.)
tfEiPS. 1,2&3
1
2
3
4929
1832
565
6605
2153
467
9333
1845
350
6267
1250
241
7480
2119
614
6485
1927
566
LIN:
69b
374
399
273
r)6)
615
L'0LA
|".
20,UUU HjUrm
L.C.LC)
VAL-(
.'ro
161
173
191
idd
193
191
169
166
167
165
1 5
163
- 1AoL) EA
i
3UUuU HJUn
L-C.L.
C)
Vi\L.(
1 '14
16'
69
16 4
131
179
165
i63
166
162
160
177
INi U LUA
"l.'
fIaj
40AU0 HJUm
L.C.LG)
VML(
170
160
153
179
1t36
163
161
178
177
175
175
172
LIST1ING NUMBErK
Cum8LJ.AI'10.4 NU 'E
3 LJwElSI
IE.
.r MERATUNE S
AVi.
LJu.
AfF T
- (
S,EAq
tELi.
LIFE
[ij.h1.
79
SA
00O NOT MAR
"N THIS SWEET-
MAA RK 0N OW LAY SW-l 4011'
am
B AT, JOHNSONi CAMPBELLAND WALKER
-
Table Al--Coputr Readout Data, (Sheet 4) (Conitinueid)
LizSTU4G f*i4BER
dOmk,1Afi0N4
NUMBER
3LOWtSf
TEM~EATUE5 3
-.-
---
37
i8
39
40
4i)
42
83
272
273
27A
2!5,
2V6
200
190
190
-190
1101
10
40
"20
220
22
220
220
14710
9477
65d0-
1650'9
82 !
1107T
8217
4014
18479
832i
I264,
196- -1932
120064 11163
62966389
14690 At43
LOG- AVG. -Lift

-AT- MiS
MPiHRS.)
255
3
396
951 LaG.LiONLY
* AUECR~)
1,2&
a~i~
2
3
-1
T762
71
384
16779
7285
1383
94'99
5704
1924,
1:6610
-7UO01 01
MEAN
445
1471
2094
1317
1723
20706
20#000, HOUx TEMP.1N0E~(

TE&P. vAL.t di) Li.Lfi
106
188
189
418
189
076
18s
155i
182
183
188
4&3000 *UR -tEMPo.1NDEFA

TE(4P. VAL.C C) L*C&Li.
186
182
184
183
1:72
169
185
163
I3i6
1'76
-173
40000 IWUR ftEmef1DE

TE&P4 VALi(C bLe-oL.
182a
17
181lab8
167
163
182
1e0
176
172
173
168
AEG. LINE
6101
A12
f37
A428,
F
J
180
17---
-1-A
44
45
46
47
48
277
278
198
199
321
322
190
200
220
240
200
190
190
220
240
200
220
200
220
LIst'iNG NUB843
CONSINATION NUMBERI
LOWEST
EPHTRS200
TEPRAUE
1
3
220.
190
200
220
L66. AVG- LIFE

AT TEST
tEMP.(HRS.)
1
2
3
10871
6188
1491
16710
9202
184d
5520
1932
516
5520
1932
468
16943
7439
1161
12976
3791
1Q80
951 L*C.Le4NLY
VALUESCHRS.)
#TEf4PSs 1.2&3
1
2
3
10422
5468
1368
1679?
8019
1174
5392
1663
508
5358
1595
45a
16755
6772
1114
9630
4515
838
NEAN REU* LINE
%534
1919
550
508
1190
995
20*000 HOUR~ TEMP&INDEA

TEMIe. VAL*( C) L*C.L.
182
180
189
188
182
179
183
180
189
188
183
tat
30*000 HOUR~ TE-iP.NDEA

TEMP. VAL*( 0) L.#C*Lo
177
174
184
182
176
173
177
174
185
184
179
115
40#000' HVJX TEMP*1N0EX

TE.MP. VAL@,( C) L*C&L.
173
170
ISO
179
172
169
173
169
132
181
176
172
/3
=
-M!
f-E
f-T
-1
ca-
-4F=~>f
-=----
N GK
T PA ?K
N'T
A P, KI0ON C'VrTh A Y
T-~
ONLY
HIEE
Ir
NRL RMPRT 8096

LISTING NUMBER
49
so
51
52
53
54
373
340
341
342
343
344
1
2
3
180
200
220
200
220
240
200
220
240
200
220
240
200
220
240
200
220
240
1.
2
8554
3262
19009
3269
6228
1505
4761
1757
12471
3269
640.
2765
1092
8437
2945
1062
3085
15447
3757
57
720
49!7
1754
570
598
4635
1633
582
1085
11833
3386
3028
917
6027
2372
570
1134
988
649
615
1060
973
20.000 HOUR TEMP 1 WDEA

TEMP VAL*(
T
C) L.C*L0
166
164
19
196
179
175
176
174
193
312
0O
176
30.000 HOUR TEMP&IN0EA

TEP* VALe(
H
C) LCL
359
1
35
73
16
170
167
17
386
173
368
40.000 HUX TEf-INDEA

TEP'* VAL*( C) L-CLo
155
153
169
163
165
163
163
182
COMBIN'ATION NJI4BER
3 LOWEST
TEST
TE&'1IERATURES
LOG. AVG. LIFE
AT TEST
TEf~e9(34R~.)
3
95% LeCoL.oNLY
1
VALUESCI4RS&)
2
OTE1kSTl
U2B3
E3
fMIEA44N
HET
LINE
193
57
1
190
187
__
_-
168
163
F...
55
56
57
58
59
60
303
305
306
307
308
309
1260
2
3
1
280
300
4544
260
280
300
12559
260
280
300
12559
260
280
300
9956
260
'280
300
6573
260
280
300
3109
2
3
1836
528
2140
720
2464
720
12*12
52d
912
529
1540
720
95% LaCeL..dLY
VALtJES(HRS.)
OTL&iPSo 1*243
1
2
3
4237
1525
496
10065
2482
575
1121,5
2658
64a
6590
1675
347
3933
1266
313
308 7
1472
716
MEAN REG* LINE
567
652
684
439
419
734
20*000 HOUR TEMP *IN DE

TEMP, VAL*( C) L.C#Le
237
231
253
250
253
252
249
244
241
231
217
21
rEI4PINOEA
30#000 HOUR
tEM1P. VAL*( C) L*C'iL.
230
224
248
245
246
247
244
238
236
225
208
206
40p O00 HOUR tEMP -INDEX

TEMP# VAL*( C) LaC.L.
226
219
244
241
245
243
241
235
232
220
2021
20
LISTING NUMBER
comutmATIUN NUMBER
3 L.OWEST
TEST
TEMPERATURES
Li~di AVG. LIFE
AT TEST
tgmp -cHiRS)
-.
N_
94:L~
RMI
81
~.
A,
-~
=44
=44
AA
':'R
4-4
44
ONoT A#
OIF1H~
AT~~~~
M2iE
1975
9-36
TES
1OA7
25783
OU79 21533
82
14282
59
1T1
LI0S00
IHOU NTMe*
CLisTiou NUMBER
~~~~~
61b
173
63
16
630
61
250
20
20
3LAT
~ 20
16
60
~0
660
2
8
0
LSthIM
i
tMBMiNAl~'
RZ
"I
-82,
NWE18E
61
68
'69
10
72
262
263
264
2,65
266
253
200
240
2 40
200
2"0
3 LOWEST
TEST
TW4PERATUhRES
1
2
3
200
200
200
200
220
220
220
220
240
240
0 240
aLpb. AVG* LIFE

At TEST
two4P.CIIRS.)
2
3
17322
5424
2184
30343
8825
5516
12469
2772
1260
29134
9228
2856
15932
1848
624
26846
3156
t140
Lt5iNLY
9 1SS.
VALUEd(HRS.,)
1
2*
16244
5665
23t85
10451
9583
3253
28908
8794
15546
1717
18338
4076
OTEM4PS. 1#2&3
2U45
4110
953
2840
628
772
14EAN IEG- L1Nt
2122
4743
1i31
2915
649
960
20#000 HO~UR TD4P * L4DkX

TkJ4I0. VAL*( C) L&C#L&
197
196
201
204
191
l58
206
206
197
196
202
199
30 000 NOUNt' tkNiP.1iMDEK

T04P*- VAI.*( C) L#C&L.
190
18%9
195
194
185
181
R00
log
-190
1a9
194
40,000 HJ UH.I TD.ND b

TWto-~ VALa*C C) LiC*Li
1E85d
184
192
187
181
176
195
195
186
185
194
190
-K
NUI4btM
I'
-'
-
4--'---'
,
191
44
DO0 NO
MARK ON Th$S $1-EET
MAR~K ON OVER! AY SHM~ ONLY
NRL REPORT 8095
Table Al -Computer Readout Data, (Sheet 7) (Continued)

LISTING NUMBER
COMBINATION4 NUMtBER
i
i
73
74
75
76
77
78
285
286
267
2685
269
270
idjo
220
i20
200
200
220
240
3 LOWkt 1
TEST
imrvAtuis
1
2
3
200
220
4
200
200
220
240
l20
LOU* AVG. LIFE

At TEST
TEMP-CkIIIS)
I
2
3
183731
3156
11 40
93471560
478
3696609045
113071 2,772
4i41
1092
149&
11996
402,
18548
14439
952 LiCiL&WNLY
P-40
i20
240
2
3
7657
1760
34199
11586
8469
2402
14044
14234
3715
874
8244'
4725
390
3990
1024
394'
1410
MEAN R96- LtNk
1019
439
4156
1061
4772
1641
18
190
210
186
196
197
TEMP1NDE"
193
L-d.L.
100
185
182
203
203
180
178o
145
111
191
187
40P000 HOUK TEMPoINDEX

'tEMP* VAfL.C W) LoCeL.
189
186
182
179
198
198
176
1,74-
111
161
146
18
-LISTiIG NU04BER
79
fib
$1
82
53
8-4
TiMP *IN0EA
30*000 HOUR
TE-t4P.
VAL.(
3967
1848
VALUESCwits.),
@TEP5. 1,19&5
26#006 HOUR
~'
LISTING NUI4gER
19
60
81
kV7
2ig
253l
252
251,
k250,
2
3
IW
200
220
240
o20
220
240
200
22
2a40
206
220
24b
200
220
240
200
220
240
1
2
2855
3660U
6660j
1140
10735
3771
20264
3010
160
3711
71 3
J9163
TEM~iHRS.
146
55
1251
95AI L.c*L.N1~y
VALUESCI4RS.)
1
2
15198
A410
4629
1447
10636
3555
0 1761So f#2&3
MEAil REG- LINE
j
3
till
1302
379
4A8
2W~000 HOU~iR
MOWI. INDFA
Tb.iP- VAL-( C) L.d*L*
199
196
30*000 HOUR~ T4PINDEA

TEt.VAL.c C) L.C-L40#600 HaUR TEMPi-INDrA
TEMP. VAL-( W) L.C.L-
C0Mai14AtZIe
NUMBER
3 LOWS
T~PRTRS
TEPRTRS
LOUG AVG. LIFE
AT TksT
'8
99
lj
1jl!
12035
4264
127.
4441_
2S871
1243
1283
112)'
1509
1423
1635
1229
1289
182
176
190
189
197
190
195
191
17
176
193
190
177
170
183
182
19)
182
1ag
184
171
1-6s
190,
146
173
165
119
178
187
177
183
179
165
163
674
JM-
V-
83
n"
WC
' --
'
.~
p4
Table Al -Computer Readout Data, (Sheet 8) (Continued)

Gomd1t'ATIj
.'J0mbEK
3 LOWE.ir
r~sr
rLiVEK
*UKE6
1
2
249
3!)6
3 to7
3!36
359
232
200
26U
260
26U
260
200
20
240
2d0
30U
26 0
3U0
26U
300
213
220
30U
240
29826
AVU. LIFE
AT TES1
T~e-(k.3
759 6
1 524
624
8 553
6439
4009
69 66
1356
52$
96 d
2952
1236
2364
1200
13544
8435
4509
9'3 L.C.L-ONLY
VAL UES(Hkb - )
%)TEMPS- I.,2&3
1
2
3
25228
9442
3379
)534
1Sd2
Aid
s961
1789
4b 9
8919
3156
1112
6948
2712
971
13375
7781
4469
MEAN' REG-. LINKJ
3717
476
560
1161
1094
4666
206
2U5
244
240
245
241
247
245
243
238
188
186
30A000 HOUK~ TEMPIkN.U~EX

TEMP-. VAL-( C)
L.C.L.
199
197
239
234
239
235
240
238
236
231
176
172
4UPUOU HOUK, TEMe IN'J0A

fE~kj
VAL.-C C)
L.C.L.
193
191
235
23U
235
230
23h
233
231
225
167
163
LOUG.
20P000 H~uK
I'Lli.
VAL-( C)
mme~. IN DEX
L.C.L.
91
92
93
94
95
96
233
234
235
236
295
302
1200
2
3
220
240
.200
220
240
20U)
220
240
200
220
240
200O
220
240
200
220
240
1
2
3
6767
3767
1169
19535
5531
1844
25351
5186
920
13532
3767
1088
5464
4211
2154
10960
4467
1848
vALIJe5(mi4Rs.
OTEMPS* 1#203
1
2
3
6457
9876
1129
19027
5627
1795
249048
4630
909
13470
3667
1084
8445
4157
2181
10919
4363
1843
MlEAN KiEG* LINE
1310
1823
965
1100
2189
1866
20#000 HO5UR TE(4P*INOEX

TEMP. VAL*( CW L*C*L*
ISO
173
199
199
203
203
194
194
178
178
188
188
30#000 HOWUR TEMP#INOEK

TEMP. VAL*.( C) L.CoL.
172
164
193
193
199
198
159
Is8
168
168
ISO
ISO
40.000 HOUR) TEMe*INDEA
167
189
195
155
161
175
158
10 1
195
184
16)
174
LISTING NU14BER
COKS3ItATION NUMBER
3 Lg-iST
TEST
T1 ATURES
__
LOGo AVG* LIFE

A? TEST
4TE1MF.CHS*)
951 y,,C.L.SNLY
TEP. 'VAL*(
W)
L*C*L.
84
lx
aw
t -.
'Inv
unZj
-E5-
NRL REPORT 8095

LI sI'F1G,&q h3E r
COINLATIO'N NUMBER
97
96
99
100
101
102
06
24
19
01
016
1 60
180
IS0
200
220
160
180
200
3 LdWESTr
TESf
TEMPERATUKES
2
3
160
180
2U0
1 60
Itso
200
P00
180
200
220
LOG- AVU. LIFE

AT TEST
TEM.(HIXS. )
I
2
3
$040
2614
793
4407
1341
430
t536
2424
540
27432
5280
1714
11573
46 35
648
182 76
4365
2568
95g L.C.Ld
VALUE(NHX
L
)
1
2
4796
2023
4228'
525d
1762
23399
5853
I0660
1297
2887
13362
536U
dTEMI S-. Ij2&3
764
413
521
1457
611
1864
683
434
61$
1597
806
2229
TEMP. L.DEA
136
136
142
183
176
156
L.C.L.
129
136
137
182
170
1l
'rEm1,4
'
DEx
129
132
137
121
13U
130
178
177
171
164
149
L.C.L.
124
116
128
126
133
126
174
173
167
160
144
137
EjAN
REU.
2U*0UY
TiMP.
LINE
HOUk
VAL.(
C)
30,000
'
HUR
TEP-
VAL.(
C)
40,000 HOUR
TEMP9.1NDE
TEM-',
VAL*( C) L.CoL,
143
LIsrING NOIMBER
C U*IJNA
NUMBER
103
1U4
105
106
107
10
66
67
69
90
91
271
3 L 0.-"4
fE f
rEMiP ERAT UES
I
2
3
240
260
260
240
2 60
260
24U
260
2?i
240
260
30U
240
260
260
160
160
200
LJG. AVG. LIFE

AT 'Esr
I
2
12192
3662
14.72
4335
12626
4662
29-939
2346
12d40
tl 76
44107
171 t39
1247
1 6U6
1524
626
1774
7t500
95. L.C.L.
,L
VAL UES(-HnS. )
,
lk'S. 1,92&3
1
2
3
11710
3669
1198
1324U
44 -0
15U
1239P 6
4266
1476
21775
2277
797
IP553
4667
1742
43454
1"327
7366
,iEA:'J
1236
1 b76
1 5"14
833
1648
7450
20.OO0 HUvc " l'E, I INULEA

]'MV- VAL. ( C)
L.;.L.
232
231
234
233
233
231
242
241
233
231
171
1,17
30,00
HoUr(
TEMP*- I NEA
TINHi.
VAL.(
C)
L.C.L.
226
22!
227
226
2;2 6
224
236
235
225
223
1 66
168
4UO00 HoUt
FMPI1.
.- D EA
VAL.( C)
L-C.Lo
221
22U
223
221
221
219
?31
230
220
216
162
162
1,1,0CHS0)
kih6.
LINE
85
'77
MARK ON 0 VE FZLAY SHEET
9,K0NTH1S 5r
0; T
NY
BRANCATO, JOHNSON, CAMPEELL ANID WALKER

Table Al-Computer Readout Data, (Sheet 10) (Coniinued)
NUMBER
4LISTFING
NtL1BEf
4COMBINATION
3LOWEST
TEMPEIkATUiIES
1
2
Tmes~c~kRs')
* eqAM MtEG- LINE
20#000 HOUR
TEP
VAL#(C
110
Ill
112
113
114
206
207
208
76
is
294
260
26
26
60
20
30
20
20
6386
2738
711
30
LOG* AVG* '.IFE

At TEST
95X L@C.LaONtLY
ALEARO
* TIP.1i2&3
109
5921
1524
672
5921
2016
672
6928
1824
52?
'5631
1606
504
13420
6652
2615
1
21872
3
5695
5864
6711
5465
13069
646
1935
669
1805
513
1595
489
5389
2560
6156
2156
692
!3
660
6684
527
503
2755
786
TEAP*I NOEX
240
240
245
241
232
223
L.CsL.
239
239
244
240
230
219
233
233
234
2!33
240
e39
235
234
223
220
216
212
229
228
229
229
236
23m
217
212
235
23
214
207
30#000 HOSUR TEkP*PINDEX

U1'. VAL*'( C) L*C*L.#
'40,000 HOUR TE.9~.NDE3
tXP* VAL*( C) L*C.L*
~-LISTING NUM*BER
COMBINAT1IIN NUMIBER
115
279
116
280
117
281
118
282
119
289
120
301
3 LOWEST
TEST
TEMPERATURES
1
2
3
200
220
240
200
220
240
200
220
240
200
220
240
200
220
240
200
220
240
LOU* AVG. LIFE

AT TEST
TEIMk'.HRS9)
1
2
3
10895
3072
1260
24313
10986
2460
11500
1848
766
28115
5448
6P.4
40682
8812
6815
14927
3760
1428
95% LeCoLofaNLY
VALUESCI4RS*)
#TEMPS. 1*2&3
1
2
3
9758
3307
1126
23113
7936
2374
8314
2290
550
26922
4124
605
25683
11999
4260
13226
4078
1262
LINE
1199
2823
663
702
5535
1352
20&000 HOUR TENP.INDEA

TkEtP* VALoC C) LoC....
189
187
205
203
191
187
204
203
211
207
195
193
30&000 HIOUR TEMPoLNDEA

TEMP# VAL&( C) L*C.L.
183
181
199
195
186
181
200
199
202
196
189
187
178
194
182
198
196
184
176
190
177
196
188
182I
MEAN
itEIQ
40&000 HOUR
TEMt'.INDEA
TEMP. VAL*( C) L6C.La
NaA-
va
*-~
wit2
'J-5
-4
411
00
iu,
qNOlT
MARK ON
~HEET
S'
M A(
O0N OV RLAY
__
fIEET OIP'
-,-=
NRL
iiiQii8695
LISTIs4a NUM1BER
121
122
123
124
125
126
P98.
299
3AS
346
347
3 48
1
2
140
160
140
160
180
200
180
200
160
200
10
200
TEMPERATURES
ISO
180
220
220
220
220
LOGo AVG. LIFE

AT TEST
TEMP,,,CHNS)
I
2
3
39040
6207
2016
45020
6212
1847
46304
9995
1421
10981
5285
1589
19970
3773
1253
33091
2765
1001
95% L.C.L.ONLY
VALUES(U HRS)
OTEMeS. I,2&3
t
2
3
31299
7200
1608
34948
7246
1426
44510
7924
1-383
10600
4216
!551
16766
4227
104$
19974
3881
597
MEAN~ iR.Go LINE
1823
1658
1567
1749
1159
796
20,000 HOUR
TP.1~INDEX
147
14o
190
170
179
83
LoC&L.
146
147
18O
167
178
1H0
142
141
144
142
186
185
163
159
174
02
179
175
139
137
140
138
183
181
158
153
170
68
176
372
137-
13
lot---
15
COM4BINATIOJN N~UMBER
3 LO-WEST
TEST
TEIPe VAL.#C C)
30#000 HOUR TW*.1IN4DEX

TE$'P. VAL.( C) L.C#L,
.--
40#000 HOUtR TEMP.,#IE

,TEMP.
VAL., C)
LC.Lo
0TN UII
COM4BINATION NUMBER
128
129
1350
01
132
'349
1650
351'
352
353
354
3 LOWEST
TEST
TEMPERAtURES
18Q
200
200
200
200
2
3
200
220
240
260
220
240
220
240
220
240
20i
220
240
LOG. AVG. LIFE

AT ftiT
1
2
35107
5472
11302
2532
3988
1512
6991
2520
11302
4303
130
2520
1253
1535
720o
816
2280 -1020F
1
2
3
31989
5827
1139
10449
2663
i409
36"2
1566
664
6905
2327
C0O
10?01
4607
2055
9336
2865
839
1461
700
844
2177
936
TM.iiS)
95% L&C&L60NLY
.VAL.UES(HRS.)
TEMiPS. 1i2&3
-.
* MEANI REG.
*40#000
127
LIN~E
311-201
20k000 HOUIR TE~esdNDEA

TENP. VAL..( C) L.C*L.
186
185
185
183.
1'.
1.4
183
82
186
184
190
188
30#000 HOUR TEMPoINDEA

* TEMP. VAL'#( C) L&C&L&
tat
I81
175
173
160
156
177
115s
178
175
184
181
HOUR TEIMP INDEX

TEMP. VAL*( C) LiC*L*
178
177
168
1.66
155
151
172
171
172
169
1S0
177
16-
i'
'_ _
. .-..
_ _
..
-'
"'
"
. .
. . ."
. ..
-:;
-A
'A
=1
jt
M
(
D~O NOT ifARKCW THISLA
I&
SSH'7EET,'
L ON
ONLY

LItrINU NUMBER
COMBINATION' NUMBEH
133
134
135
136
137
138
355
325
32~1
329
330
371
3 LOWEST
TEST
TEIPERATURES
1
2
3
200
220
240
220
240
260
220
240
260
220
2"40
260
220
240
260
200
220
240
LOU. AVG. LIFE

AT TEST
.EmPeicmRS.)
1
2
3
19325
475
2364
10512
7553
2604
2520
1t6o
766"
3773
1788
151-2
3773
2772
1428
1'4090
2520
624
951 L.C.L.0NLY
VALU9S(HRS.)
OTEMPS. 1#2&3
1
2
3
15788
5698
1923
10000
5470
2513
2345
1303
713
3063
2056
'1222
3680
2360
1403
13054
2652
577
MEAN MEG.
2157
298 5
747
1376
15"8
603
20D002 H i r.UduR.
TEMP. VAL-.( 0)
.C.,L.
196.
195
207
197
163
157
158
13"7
165
153
196.
195'
30,000 t.OIJR TEMP.INDEX

TEMP. VAL*( C) LiCaLe
191
185
197
15
153
147
146
123
153
139
1!3
190
186
13-
190
176-
1'46
140
138
113
'145
130
t88
187
LINE
40,000 HOUR TEM0.,NDEX

t TEMP. VALs Cd), L.C*L.
4-
F-
P-N M
LISTI4N~UMBER
13i9
140
CQBtAT0
UME
72
374
1115
200
220
-240
200
'280
240
200.,
260
240
220
240 '300
141A
tdi
283
914'
01462
i45
838
2682
660-4
838
11343 4077
1649~ 162
62t4
720
7681L
925
8;16
952 L..LiNLY
VA~UES i4RSO)
OTEMPS. 1,2A3
1
2
14367
3430
27302'
W9~2
!it3
l1458
2569
345
$09
8344
2019
45A
4171
i662
719
6720
2212
715
AMAN RtS. LINk
872
564
957
543
722
770
8000 EM.IbE
4LJ
TEMP. VAL-( CW
Lo.CiL.
18
195
200
192
205
204
251
248
2i9
229
244
241
30i0QG HOUt TW*PINDEX

TEMP. VAL&( CW Lk#CdLo
193
190
195
187
201
199
246
24 ,
221
2il
237
235
40.000 ROU8 TEMP*INDFA

VAL*( W) L.(.Lo
190
186
192
183
198
196
243
238
216
916
233
230
3 L8WEST
TY5T
__TEMPEI4TURE.$
Lid'.V~ LIFE
AT TESt
-TEMPi
1
3
1
88
-2
Pat-,
F'7
La
142
I4
143
14A..
50~
364
365
2",
260
2$0
30'0
ii
300
"
41-
TET9~
'2
4W
- 24350
LL'G~JCDR14
3L3gS
AT
iToo:NT
TtP.~
.
..
IM
i
3
viu
VALC 0l) L iC.L.

40i00)4UR
TEP.e4EX
VA
aop
109 --.6655
2444050
-4E
tl
TIMP.
406
32LI
M TA$5TEG
40I ~u~
14~
26
156O3T 127095ll
47s
1960'
100
202
20.1
270
4
-,'141
4Ili
1960
21-!
142
191
:23
i
87
-C
1-
050-
26 65
,489
1206
F4
-93
24
109
84 48
191
6172s
--1137-
1179
232160
A'-i
176
840
117
1903
18
ji
-172.
--
4-66
-14
--
L1ST1I4G NUM*BER
COMB8INATION NUMBER
,4~.'--951
OfA
151
152
153
15-4
1,55
156,
242
338
290
291
1ds
16
3LOWEST
tkST
TEERT E
I
2
3
200
260
240
180
200
220
260
280
300
260
940
300
260'
280
300
260
2806
300
LOGo AV~o LIFE

AT FE5T
TEMP.CI4RS*)
1
2
3
I15680
7383
69
10512
3573
2177
22600
3317
33
22600
6,1 d3
i926
3612
1248
526
4608
1416
622
LoCoL,6O4LY
VALU95($RS.)
ifTEMPSi 1&2&3
1
2
3
12.972
8222
4659
8624,
408*1
1778
15861
4191
938
2iA76
7901
2916
3433
1291
501
4078
1517
550
ftLG. LINE
5231
1990
1147
29
516
592
20&000 HOUR TE~e*WNEA

LeCeL.
TEAO. VAL&C d)
189
tat1
164
159s
260
257
262
262
224
227
233
230
30#000 HOUR TFJ4P*INOEA

TEI4P# VAL*( C) L.C*Li
175
165
155
149
254
251
255
255
222
220
227
223
TEMP*INOEX
40o000 HOURF
TEMP. VAL*( C) L.C*L.
66
154
i-49
142
251
246
250
250
217
215
222.
218
1
~-~~-~~
1
Jar9.
~-
r-fa~-~--
EEL.
k-All
rN~-
4iLi
7M
7
;I-SI WM-tt8Im
d___
V
~
il-
__446_____
CAPL
tohiAi
A
00
ZMLATR4
-5
__
_A
iIN
____N _4,,,_E,
___
____
-~_
-11
spaV5-x:3r
-4
4if
MDWAKE
_4i16
1600
300-
-:00,
300
20
-1237-
-863-
-15S
70836
TE
ir23-
24O
148
72
591-~L.0L- NLY
;U
39T6
258
_152-
jt435,
4T8MP5.
1.283
700
3-23-
574
15
,-
2471459
9
-1,92
421
-624
-"34 5
_6
-4
ljhLI
VAN
4,!
1~
TENP.(M[St-)*-
3"' 01
201
5034
3524
T"
'p
L"t iaN 4~ UME I
RAT ES
__1DP
L0.AV.LL
~ 1
2
3
1,4"t4ii
ATtEs't
SZ..L$
MtAN HICU- LINEt
.26004 HOUR~ tE~iMPNDE

WkO- VALi(C ) L ..
TE~iP
3000
HUSTEP.lDX
L-a.L,
A.
4dko00 i4OUA- fEMP.iJEX

EEMP'i- -VA~.. W LiCiL
...
.
Jl_0
a_2
1'A
240
240
g40
240
4
431 5
i612
Q78894
7589
2161
45#0
1260
648
28455
6937
4463
44 6
7410
i443
8s
1600
21833
091
3821
1354
1004
194
14i
205
205
1,72
166
904
200
200
182
179
8
i6
fi
1949
165
17
li
19
i59
194
146
19u0 11
183
81
19
95
L61
191
191
111~
154
la3
186
1 66
240
24
2i70i'
7919
2259
1083
~~i
1573
6.5
107
-4A
171
go!
vvi
---
1--
~57~
______~
4.
Q4
AZ
-AM__________
I;
_____________________________
________________________
_____________
to,-'
NRLWRT 8095
Table AV-Comi utei-Readout-Dat, (Sheet 15) (Continued)

M13ER
____'OM1AL
3LOWT-
TEMPEATUES
t12
13
-14J,
tO7,7
i45
1200
'V-
2A
20
240
3'2
2_
20
20
83
_238,
844V
31-88
_69 2
1848,
24315
-8469
311'
20S09- .46
2005
~0
9
40
AI
AT TEST
EmI
_2
3-
iCHRSil)
9~ZL.CL.~-
~5780-
-~-VLEC~~
4A
-424
A-926P-1:749
'5034
-2856
153
974
'-61:65- -510
099t
2717
-1~~
i243,
-12
G.LINE
-3
15Ml49
2000 HOUR
TEMPLIE
24
aooo E~i~~x239
II~R
Tkt4.AL
----
q)
l#.
40s000'HOUR JEt1-_d4oqEx
VAL.C C0 ~ e.
-~TEAR-.
2-
23I
35
226
23,6
_2'6,
31~20
21
11
431-
-6041
3181V
149-
244
241
23201
303
0
f
208,
236,
234.
192
263_
20
2351
3
2312
195189
1 75
176
177
lid
140
141
154
15
240
260
240
260
n220
240
260
8692
1116
5712
840
616
8676
5080
252
491
VTEI4PSv 1#9&3
1
2
3
157
3545
795
7358
1749
224
M~EAN REG- LINE
807
20.000 HOUR TEMP *INDEA

TE#4P* VAL#( W) L,.CoLo
LISUA~G N'UMBEK
COMSINATZ0O4NUMatI1SR
5~
1I0O
157
'300
260
280
300
Ai2
6474
251
1055
6474
2518
1008
5755
1522
326
3715
1121
271
6305
2504
1028
6456
2478
1006'-_
396
416
346
1053
'1015
197
197
216
205
207 .201
193
202
238
237
238
238
30.000 HOUR TEW4P.NOFA

L.C-Le
TEMIP- VAL-C CW
192
192
211
00
203
196
195
187
230
229
231
231
40s000 HOUR TEI4PdINDEA

TEIIP. VAL.C C) Led.L.
189
1a8
209
196.
199
192
192
183
225
224
226
226
3 LO~WEST
TEST
TEMP~ERATUJRES
LOO-0
-~AT
~2
110
AVG- LIFE
TEST
____tEM~i(HRS.
961 LoCoL.0O4LY
AVALUtSCHRt-)
200
220
24U
1
2
3
15983
3288
1
-2
3
220
260
-280
wv4-
5:j
EEL
4
-w
.4~.~4~44
91
'AA
W-3S
MMM
ILI
k4
Ib
----
-66
(A
i-
jot
41
14
66
4-,--
-'I5A-241
427
COMSNATIG
NUMBER
15
AT TEST
TEMP0--.$.
I
29
'
36
31904
212
'285
543
35i4
23
115
100
2143
-233-
785
-VAL-tSCHR-oI
2734
24 1
2237
553 2
3000
TDP440A
0lU
20
234-
203
2311765
3L0aL.
<'
219
TEMP.-, VAL~wC C)
&~
dAD WALKERj-,
15916
-
ti!A97f4
5--9t'3
incAoJO
25N 8A66EE
_______
20
24
62
95
bd
A0g
19
2&5
563
732I
a
----
=-
198)
--
1'7 -706
14511e0, NUM4BER
C#RNTNMM
1s?
led
19
191
192~
97
1og
109
57
84
85
2QO
20
24
200
220
40.
200
220
240
160
180
200
2,40
260
280
240
'260
280
6690
i!84
456
141123 1192
382
15381)
346
120
11040
86,40
4601
9229
6215
2762
19271
6105
2'430
13327
2312
366
14139
1383
133
10711
7129
19341
6900
2389
17913
6313
-3 LOWEST
fiST
tfCPEATUJRES
t
*2
3A
LOG&0. AVG* LIFE

AT TEST
T14.HS)
9kZ Li:c'L.0NLY
9.
VtJSHS)
t,'2&3
EMPS*-----TD~
1
21818
6261
435
4311,
2257
II
HO~UR TEOP-IN0EX
TM.VALiC C) L.C.Lo
188
182
198
195
200
197
140
130
241
239
239
238
TEMhe.INDEF(
30,000 1i8Ui
TEMI' VAL*( C) LoC.Lo
182
176
194
191
97
194
26
113
34
232
232
231
111
1881
191
102
226
225
~ii00
TEMP. VAL*(
C)
LoC.La,
Qffl
--
Rx~
____
j4
7".
sm
LLT1
S -O
3 -
0-0
St
60I
I
L
N_
___
TEKFS-
25
-68
3-
-142
45l REOR
NUM8ER
-Idow -kAiJN
16i
-W_5____
-,31
LT-IIE
gd
6A
i,7,
9,51
48
19
19
14
llliE,491
84
1'
-24
809546
4931
1948
92
366A
9
368a
U6
37q
-0
367
4553
4I,,
-6
LLNU4~I19
201
202,
203
0A
3 kWEST
1-
160
2040
200
200
P.00
200-
TUMPEUAfUIRE
160
240
240
i240
240
240
L@Q. AV4. LAft

At UEST
fiPCR.,
9SA Li0.L0NLY
1
a
3
20922 15557
91"
i846
542
549
19197110
18077
3646
76
14045
2i514
756
196
6384
336
168
2S1
6027
*56
1'"
51
*EI.'23
392
14"N AEGi LIN~'E
__
200-o
403
5137
~864.
3703
27810.
577
916
839zA0
642
74
156
A79
862
705
111
169
2b000 HO0UR TEt4P-INDEJ(

TEMP. VALoC W) L.O.L.
160
-1
196
193
199
199
195
194
186
176,
188
18
400000 HOUR 46W o 114DEX

TtEMP. VAL-( CY L.C.Lo
150
146
192
4a
1%4
194
10
186
18
a2
7
184,
183-
40*000 HOUR .TEMPP.IN0FA

t640. VAL*( 0) L-G.L.
142
140
18#
165
191
190
186
1641
1-79
168
18
l8d
__
=L-~-RPM
me"-
pet
I.A
--
'
-2
NU
-'
TEPRAUEANCATO JOHSON
____~~i
'10
_240BEL
AND WALKER
,-24
Table215
C,
t1
-k
20
S-'
193
-!
- L0U.-A'VaS-L1TEMP
i3~91'
AtTES2"2436259
149
-0 96
14OBU-0-4-9,SE
-14-
1'316
______4
278
77
__0_
too
AT
'
-A6
?o- -b 20
183
Ul 3.WE1
19
49 -
1538
163
44
214535692
IO
4365j,
t81
64
_,
L14tiNG NUMBkK
00I181NAt1t6N NM~tR
212
21;
914
215
216
i 20.
121
'122
1430
131
136
200
220
240
200
220
.240
200
220.
240
180
200
220
200
220b
240
3 LOWEST
TET
TEMPER(ATURES
.3
200
220
240
L Wd -AVG-,'L1Ft
AT TEST
I TEMP.6C HAS.)
I
2
3
436fd
766
109
10299
2564
479
11349
2384
600
27656
3864
$07
15456
3040
P.I8W'
32976
4283
1104I
4264
97-57
10417
27138
14309
26523
a
3129
14
2A147
458-
2311
-551
3575
519
5308
201$f
49118
884
136
515
594
545
2109
1004
185
184
193
191
193
192
203
203
175
174
205
203
30.000 HO0UR TEMPoI0EX(

VAL#.( C) LoCiL.
181
180o
188
186
188
186
200
199
167
166
200
199
40*000 MOUR T&M F * IN0 FA

1e.vAL.c CW L4C-L-
178
17
ids
18e
185
m
i97
196
162
161
197
195
11
LGC*L*ONLY
,951
VALUE$CM(
HS.
~T~~'1.23
MEAN REG0. LINE
20#000 HOUR TEiMP&INoEX

TEMP. VAL,( C) L..C*L#
71
all
tA
A
-
IR
p
7I"MEE.
__
-
~-
==~-=~
-- ==--__
~~
ME
-20~
-A-'
II
1'37
-ij
-l
IEMSHA
--
2612
_&boo
~
ffi fm
-T
4t
4k
3.;L00ESY
AL
9 L-O.sOL-
-6
1M
52-
147'
49
101
20R;28'4
-O1
1i~L
-.72
+36 ik
50-
2695
-2
-- 2
-'
'a
83
-15
95
iI
i d --
-222
X-
34-
-5
42
i1-
567
-O15-
2071
1A3
42'66-
54
____06
LIAGf.hf
22
NUNWE
t0M8lINATWON NUM8E A
.224
us
226
227
228
14'
190
200,
0i
260
22
240
17270
08
tilt6
6216
14
364
013
01A
12
13
2009
220
.40d
3 1.8WEST
fEST
tikPiRATUIRES
I
2
3200
I
16U
ISO0
190
200
220
I80
200
20
LVGo AVG* LIFE

~AT tST
TENPoCiHR5)
1
21579
3
6014
585,
12536
i32
T69
1144i
3454
2200
5208
1694
t64
95t LiC*L*ONLY
V5U~IR.)
OTEMSo 1#'2&3
V5482
2
1680
532
'12507
S59
013
114,75
4553
1434.
11
5096
162171Q06
1471
543-
1015
1824,
562
1069
406
175
166
179
178
186
183
186
183
136
3
168
182
159
ISO~VLC
8
173
172
181
177
182
0&8
132
130
1-79OJ7
169
163
154,~ 167
177
113
178
174
AgAN REG- LINE
3
j
561
186
141
144~18
20&600 HOBUR TEMP.*I NOEX

LC#.
0
TEP.IdL.
30#000 HOUR
)
TEMP NDEX
L0L
TEMP.INQEA
40#000 HOUR1
t$.VAL-(C0) L.C-L.
95
7a~r.-'-KW-j~l
1?-N
'ZEE
ft.
5931
i500
351
f.'q
--3-
-+-M
A4I-4
Ad
44
2174
ig
-2031-6
/j3
-,
16-
OJ
-x
-Uk
tIo
at a
"
1i
~~ ~
R---,JH
~ ~
13
--
2W
22
f---3
1'4
39
C-i-*
C)
It
~SO ~,C
_'0-
1-2-
D
MPELA
-- --
Table
A1ComputRadotDt,(he
AKR
0 Cniud
LISTING NUMtkit
235
do
COhSi8LATIION NUMBER
3 L0aWiT
TkNIeAttU*~CS
.00- AVG. LIFt
AT TESt
TbW-~C($.)
LY
di
.U
9 5-L
VALUE~iCIAiS.
-O~~tE4I&.~ 1.~s3
MtEAN4 REI
LM4E
HOUR
---.---
qv.-
'N22
ye
240
017
old
02
16
li0
200
160,
180
00
160
is0
200
335
.3
dA
I
k9025
1008
1428
?id0
849
0
31 16
1504
.q
563
4
512
S2
30
148
1192
1
2
399
6664
9 580
7527
3199
1421&
168,71
7297
3066
2673
1255
503
4967
2497
1123
15s7i22
4709
1467
1066
1456
3325
599
1283
1541L
121
119
120
119
138
136
123
111
131
122
157A
156
114
112
130,
116
123
151
Ili
109'
106
111
IU7
1I
AYS
127
124
124
103
tit,
97
113
117
106
15w
147
146
TEMP.UJ0EX
~-ma
210
140~
1460
lad
180
96
-~
,0$1
238
140
i140
rib16
66~
"1
166
TtMP- VAL*( b) 'L.b.L.

116
40;000 HOiUR ~T&W'
FMe oVAL*( W) LoCi.a
44
237
I
2t
3
10UA tEtiP-.1NDE
24000
UL4IF VAL'.( C) .L.C-L-'
______30#000
236
-AM-
S-6 4
44
ftOT-89
NR
00
0-
_4
______At
4469
________
22v-
-1-400
14
-i
1o
EPs-0 AE
3C
AVG._LIFE
L0G.6141Ji
A-TTh L t
TEMP.885r
160
14i
20
A
160
23Li
221
-14O
200
0
21142
T73
94255
45;
94
2503
24
LITN IE42419
3tWS-t
- EST gU
TEiP*- 'ATURE
-5-
666
REPORT78095
a-_
_9T
2046
22
18
2007
2420
'13
200
162.
:1648
t7008
1-4909
24
I
14
-45
344
-245
_
240
145
2906
220
I2
10
175
71
2
!804
200
2208
6
63
4
LM5i
247
248
240
950'$16,5
mLmatii
-43
44
74
76
11iftIN
0W4MiNAfiii
~4
1OW
180k
190
140
140
200'
200
160
140
220d
220'
380
fld
ad0
220
240
10330
4024
1800d10.0
4199
7546
1,408
1713
-00i
97
634
371
456
404
10077
75
315
d58
1208
9
6?2
12531
74)
161
2705
2744
0972
j36
-424-
A47
MtAN NEG. L.LE
340
132
586
jq
39
605
S43
20*000 HOUR TEN3'.IND"3

TEM4 VAL*( C) -L*C.'L
'174
173
179
174
197
3124
1e9
328;
199
194
200
196
30*000 HOURTEN?.INDEA
TEMP. vALoC0 C Lo.
170
168
176
171
121
119
124
12S3
195
189
196
190
It
118
164i
i15
121
120
192
185
32
1871
TEST
7 *4p1amtUftE
L.@o
'At
AVG. AIFE
-TW
1,
W 49
318.s
60
tDP(S.
95S 4L
iv~i4io~y
11
VAU5$S.
3
----.
ta.123
46000
HOUR
tEI41'.iNE
TEMP. VAL-.( C) L.C.oL.
1-671,
165
l00
iSO-4Vj
40
15600
'35
5413.
i'4212
___
~---I-
INI
EM
--
IM
o4-v;
BRANCATO JOHNQ,
-~~
AMP
LAND WALKER
3,Li31
T~EMPE~ut~K
L94- AVU. LJ~k

IESt
-AT
4 -
sL
iik~fjso)
95% LoCiLiONLY
VALUF.S(Its4.)
G~i2s.
',~&
M1EAN~ jiEG. LINE
1.29
200
200
34
240
0-4u
I88
2sl3
10
18
51
00
46's
1
215
3
3
51A
77
t,440
b32
393
201
451
200000 H13Ui
TEM'PoINUEA
iWvP.o VAL o C) t..C.IL
193
102
184
1SO
30,060 lidiOU
TEm~p. 'onwx
ThilP* VAL*( C) L.C.L.
189"
186
119
17
000 HIOUR TEMP.:-AEX

-TE1,' VAI.&C C) L.C#Lo
187
'185
176
171
***EN'D OF
TIME~S
DATA
77.45 SECS,
#*
40.000 HOUH, TEAP.INDEX

Mtn. VA.( C) L'C.L.s
***END OF DATA
Titt
187
176
171
**
77.45 SECS.
KA
2e
--98
VG
tg
NRL REPORT 8096
I-SE
va
to Computer Data Sheets (Table A2) and Motorette,
~Index
____________
Sheet
Number
Listing
Number
23
System Magnet
Number Wire
Gud
1A
23
1A
23
15
2B
23
29
2A
41
2D
42
2E
23
95
2B
23
23
9
10
98
99
A
A
2E
2E
A23
Svstem Identification
______
v23
RY23
9Glass
23
11
Y-3
10A
23
12
12
B-2
2A
24
13
36
24
14
14
24
15
20
Z-1
2A
24
16
26
Z-1
6A
24
17
OilMica-Glass
24
18
49
Z.3
2C
21
Y-1
6A
C-5
10A
2A
99
IMI
slot
Wedg
Cambric
Fish Paper
Org. Varn.
Glass
Cambric
Org. Varn.
GMG
Mica-GlassJ1
Mylar
Film
Org. Vain.
Glass
Org. Varn.
Glass
Org. Varn.
Rope
Acetate
MylarPaper
Org. Vain.
Mica-Glass
Org. Varn.
Mica-Glass8
Org. Varn
Mica-Glass
Mylar
Mylar
Film
Film
263-3F
263-3F
Copaco Rag Copaco Rag
Paper
Paper
Sil. Varn.
Mica Mat
Glass
Glass
Org. Varn.
Org. Mica
Glass
Glass
Epoxy Var. Org. Var.
Mica-Glass
Mica-Glass
Org. Var.
Org. Var.
Glass
Mica-Glass
Epoxy Mica Epoxy Mica
Mat-Glass
Mat-Glass
Epoxy Mica Epoxy Mica
Mat-Glass
Mat-Glass
Org. Var.
Org. Var.
Mica-Glass
Sil. Var.
511. Var.
Mica-Glass Mica-Glass
GMG
GMG
GMGAl
GMG
GMG
GMG
GMG
GMG
GMG
GSG
GMG
ZI
GMG
GMG
0MG
0MG
GMG
GSG
Index to Computer Data Sheets (Table A2) and Motorette

System Identification (Continued)
______________
Sheet
IListing ISystem IMagnet I

Number
ans
Wire
Number
Number
24
19
22
Y-1
10A
24
20
33
Y-2
10A
24
21
19
B-i
2A
24
22
23
B-i
3A
24
23
32
*B-1
3A
24
24
34
B-i
2A
25
25
40
B-i
4A
25
26
30
B-3
2A
25
27
31
B-3
4D
25
28
C-2
4E
25
29
37
B-5
2A
25
30
53
16A
25
31
67
C-i
4B
25
32
68
D4-
NONE
25
33
48
C-i
4A
25
34
51
H-3
10A
35
52
H-i
13D
IR,25
*#185
AWn
100
I_
hs
______
rud
_______
Sil. Var.
Mica-Gimn
Sil. Var.
Mica-Glass
Org. Var.
Glass
Org. Var.
Glen
Org. Var.
Glass
Polyester
Fiber Mat
Org. Var.
Mica-Glass
Org. Var.
Glass
Org. Var.
Glass
Org. Var.
Glass
Copalum
(20 m)
6105C
Sil. Var.
Mica-Glass
SO. Var.
Mica-Glass
Org. Va.
Mica-Glass
Org. Var.
Mica-Glass;
Org. Va
Mica-Glans
Org. Var.
Mica-Glass
Org. Var.
Mica-Glass
Org. Var.
Mica-Glass
Org. Var.
Mica-Glas
Org. Var.
Mica-Glass,
Glass-Mica
Glass (12 m)
6105C
slot
Wedge
S
GMG
GMG
GMG
GMG
GMG
GMG
GMG
GMG
GMG
GMG
Polyester
Var. Glass
SO,. Var.
Mica-Glmn
64051
(Dupont)
Polyester Var. 0MG
Mica Mat
GSG
SOl. Var.
Mica-Glass
Org. Var.
Org. Var.
GMG
Mica-Glass
2 SOl. Mica
Glass
1 HT-1
Paper
Mica-Glass
2 Sil. Mica
Glass
1 HT-1
Paper
Silicone
Glass
Molded
Glass
NRL REPORT 8095
Index to Computer Data Sheets (Table A2) and Motorette

Sheet
Number
Listing
System
Number Number
i V25
IMagnet
jWire
Slot
PhedgdSlo
36
54
H-i
13D
26
37
59
H-i
13D
26
38
69
H-i
13B
26
39
70
H-i
8A
26
26
40
41
71
73
H-i
H-i
8A
4E
26
42
74
H-i
6B
26
43
76
H-i
10A
26
44
80
H-i
13C
APaper
26
45
84
H-i
10A
26
46
88
H-i
10A
26
47
94
H-i
10A
48
58
D-3
4A
27
49
57
D-3
10A
27
50
85
D-3
4E
27
51
52
86
82
D-3
I
4E
10A
4Z26
Al127
4V(HT-1)
(ML Var. Gi.)

6507 (7 m)
(ML Var. G1.)
6507 (7 m)
"H" Film
(7 m)
Sil. Var.
Mica-Glass
Doryl Glass
Sil. Var.
Mica-Glass
811. Var.
Mica-Glass
811. Var.
Mica-Glass
Nomex
(HT-i)
(ML Var. Gi.) 64051

6507 (15 m) (Dupont)
(ML Var. G1.) GSG
6507 (7 m)
"H" Film
GSG
(15 m)
511. var.
GSG
Mica-Glass
Doryl, Glass
GSG
Sil. Var.
GSG
Mica-Glass
511. Var.
GSG
Mica-Glass
SiI. Var.
GSG
Mica-Glass
Nomex
GSG
(Htf-1)
Paper
Nomex-H
Film-Nomex
Nomex
(HT-i)
Astrotherm
240-21
Nomex-H
Film-Nomex
Nornex
(HT-1)
Astrotherm
240-21
GSG
511. Var.
Mica-Glans
511. Var.
Mica-Glass
Mylar xm633 (2 ply)
633 (1 ply)
Nomex
511. Var.
Mica-Glass
8SR. Var.
Mica-Glass
Mylar xm633 (2 ply)
633 (1 ply)
Nomex
(HT-1)
Paper
GSG
Paper
GSG
GSG
GMG
GSG
GSG
GSG
101
~1

Index to Computer Data Sheets (Table A2) and Motorotte
Sheet
Number
Listing
Number
27
53
100
L-1
NONE
Nomex
(7 m)
27
64
12
F-2
10A
27
55
87
F-1
UE
27
56
58
D-2
10A
27
57
60
D3-2
4E
27
58
75
D-2
4E
59
77
D-2
4E
60
78
D-2
SC
811. Var.
Sil. Var.
Mica-Glass
Mica-Glass
Cfamn 150?POfam 150?P
im/5mI5m 3m/5m/3m
S11. Var.
Sil. Var.
Mica-Glass Mica-Glas
Dac.-Mylar- Dac.-MylarDac.
Dac.
Polyester
Polyester
Mat 2542
Mat 2542
Estermat
Estermat.
Dm 70-353 Dm 70-353
Estermat
Estermat
Dm 70-353 Dm 70-353
Estermat
Estermat
Dm 70-353 Dm 70-353
Duroid
Duroid
2307
2307
2'27
27
System Magnet Varnish

Number Wire
28
61
79
D-2
SD
28
62
81
D-2
4E
102
PhsGoudlt
Wedge
Nomex.
(14 m)
GSG
GSG
GSG
GSGA
GSG
GSG
4
CSG
GSG
1
GSG
Duroid
2310
'Off
Z4
-A
-I-A
MARL
'OT MA~f( N THIS SHE
VRL
NRL REPORfT 606
T~ble A2---Cbmpute R,6adout Ditta, (Sheet 23)

~
-3
1~5
29
4l
I,
21-60
35
3-50
-160
IS0
200
145
160
1S0
1,59
160
'145
160b
180
135
1-6
1180
ISO
'Ido-
It
2
2118
156
I
83642
25ib
S
2A0
914
460
3926
540
31-70
21
2418
945,
821
289
3860
$24
129-
863
113-
1366
522
d3g
046
3i2
848
179
587
113
13 10
130)
109
103
!qsYTEhM NUMBER
'LOESt
TEST
tEilsFTU4ES
L~gd) AVG- LikF
At TEST
--
19
-VALUESC NRI
R
'TEMP5.
O
W'2&3
OikAN REG. 11.4k

--
1573
2,04~~
21
1533
93.
-120
126
_!EM$.- -VAL.C0)L.CL
TF.MF-.NDEk
091
119
0.
00'H0IJ
4
TEAP~ V--ALC)
TEfV',1NDEA
kCL
d4
113
120
12
16
312
DP~1NDEA
EM
79-
I109
4000 f
AA
I--
-1-16
405
Ii
10
1 6
li
9
101
123-
968
LISTING NUMBER
SYSTEM NUMBiR
42
3LWS1.145
2
TEST
.AT
9'
140
160
S
18010
140
160
192-
240
260
20
160
I80
0
4760
tauo
too
200
ISO
1
2
3'
3689
1722
921
5600
3058
1606
3236
1249
479
6500
2184
i020
2274
680
25V
95Z LoCeLoONLY
VALUESCI4RS.)
#TEMP~S* 1&2&3
1
2
3
3123
1812
781
5486
2925
1577
3173
1198
471
5921
2325
927
2096
74
239
4603
16
697
LIN~E
883
1633
487
97
25
73
20#000 HOIUR TEtlh*Ie4DEJ

TEMP. VALC C) LoC.L.
108
99
106
104
125
126
118
116
20!)
203
134
132
30p 000 -NUN~ TEMP.-I WDEA

TEmI0. vAL.t d) L..c.L,
100
90
97
93
121
120
111
4109
199
196
127 j
125
40*000 HOUUR TEMP.i.'DEX

TEMP. VAL*( W) L*C.L.
95
84
90
67
117
115
106
104
195
192
122
120
LO0G* AVG&. LIFE

TkSt
_______s-
MEAN NEU*
103
mAA
to~
UMIgATUR1ES
9~r
1
t796
720
II
T
hAARK Oi
C4
0:
N
~N1 A KVU
fA15
Table A2-Computer Readout Data, (Sheet 24) (Continued)

LISINGi
13
14
36
14
190
200
22b
160
IS0
200
I1
2
3
16500
5342
915
5680
2200
552
05%-L*C&L.IINLY
VALUES(HRS.)
ST&PS 1,2&3
1
2
3
1-4510
5594
Al2
* iEAD4 REG* LINE
vUMBER
1S1
20
26
49
21
-160
180
2k0
60
10
200
160
IO
180
200
200
22i40
5302
2790
1245
6965
3793
2400
8256
1478
5855
5440
1400
382
5484
1766
539
5058
9513
1194
6651
3686
2290
622
1745
439
.5421
1372
381
890
606
1296
2357
520
385
:0,OO0 HOUR TMP.INDE

TEMP. VAL( C)' L.C.'.,.
188
187
142
1-39
130
124
126
124
147
142
183
182
30#000 HOURt TMP. NDE

L',G.L*
IMP* VAL.( .C
184
183
136
132-
121
115
115
112
142
137
178
117
TEMP.1iNOJ
40,000 HOU
... VAL*C ) L.CaL,
181
ISO
132
128
115
10s
10
104
138133
174,
LISTING OUMBER
19
20
21
22
23,
24
SYSTEM NUMBER
22
33
19
23
32
34
200
200
-- S0
ISO
160
160
220
-220,,
200
200
-.160
too
SYSTEM NUMBER
3 LAWEST
TEST
TEMP~ERATURES
7,LO8. AVG, LIFE
ATTEST
TMP-($RS.)
1
2
c3
3 LOWEST
AgST
_.
"174
p-
23
2A
19-
23
32
34
t80
20
220
4126
1750
141
I80
0
220
6400
2070
774'
160
10200
6934
53j0
2900
160
10
200
16997
8530
140
L14TING NUMER
19
20
21
SYSTEM NUMBER
22
33
200
220
240
2
3,
167o
310
200
220
240
5666
1130
26
05% L..G.1-.0NLY
VALJESCHRS.)
# TEMi$So 1#293
1-6046
2
3
1368
30
5341
1120
248
4058
1.69
730
626d
2100
758
6674
4538
2814
15501
5138
1317
M4EAN R90- LINE
337
262
751,
767
3104
1735
aO0*000 HOUR TEM~IP&NDEA

TEMP. VAL*.( W) L.C*Lo
187
185
185
184
149
147
161
160
122
168
160
155
OUR TEe,- '~bE

30,060
TEMP. VALiC d)->L#CoL.
182
180
181
IS0
141
140
154
154
16a
9
154
4
179
176
1.78
176
13
15
'150
150
go
at
150
143
1
2
3
3 LOaWEST
TEST
TEMPER4ATURES
16280
AT0TS
TEMP oCHiR$.)
46,6
HOU
iTi'r. VAL.
TiM14 INDEX
C) 'L*C**L
104
-i
Ile
A
NO2 K0, II S-~rj
AA,
-4
Al
Py
NRL REPORT 8095
Table A2-Coniputer Readout Data, (Sheet 25) (Continued)

.LISTING NtA4BER
S-----YSTEM1
7-
NUMBER
25
26
27
28
29
30
40
30
31
39
37?
53
3 Lowest
TEST
TEMPERATURES
I
2
3
180
200
220
160
I80
200
160
180
200
IS0
200
220
180
200
220
140
160
ISO
L(4Gs AV~3. LIFE

At TEST
TtMP*CHRS#)
1
2
3
7687I
3245
953
13368
5064
2676
22104
7211
31,45
7791
353U
540
6480
4540
1129
8 427
730
430
95% L9C#L*WNLY
VALUESCi4IRS.
#TEM4PS. IP2&3
1
2
3
7)82
2665
877
12049
5410
2407
18842
7392
2682
6866
2126
48tS
5955
287o
1059
4252
1153
214.
M)EA0' REG* LINE
1027
2557
3035
665
1369
316
20,00Q HOUR TE.I4DEX

TEMO. VAL.( W) L..C-L.
165
160
150
146
161
159
170
162
161
147
127
117
3Q,000 HOUR TEMP&diOEA

TWO*P VAL*( 0)
L#C*Lo
15?
153
142
139
154
150
185
155
153
137
122
]111
40#000 HOUR TEMP-1I'DEA

"~~l.VAL-0i W L..C-L.
154
148
136
133
149
145
163
151
148
130
119
107
31
32
33
34
35
36
67
68
51
'52
54
LISTING NiRIBER
,YSTDI
NUMBER
.. MwLOG*
---
LISTING NUMaER
31
32
33
34
35
36
SYSTEM4 NUMIBERI
67
68
48
51
52
54
200
220,
240
200
220
240
260
*260
300
240
260
280
200
220
240
AVG*, LIFE
At AS8T
tmCikkit.)
1
2
3
89,6
goo
209
501 7
1137
312
9655
19$5
5 *6
10093
4046
8570
9538
3049
1196
350
2100
75
951 LeCeLeON4LY
VALUESCHftSe)
YTiMPS. 1*2&3
1
2
3
6294
1017
146
4867
.1160
302,
88 38
2106
526
?,565
3015
835
8899
3124
1116
3380
16734
746
MEAN lIEGs LIN4E
179~ 30d
553
984
1170
842
1:
280
280
g29.000 HOUR~ TEalP*cNDEX

TEMP. VAL*( C) L.C.L.
191
1ad
1og
182
190
189
251
241
227
225
202
193
30,000 HOURI TEL4P.iNDEA

-TEMP., VALCt W
Ldo.L-
187
184
177
177
1m8
184-
246
241,
220
218
194
184q
40.000 HO8UR TEMP#INDEX

TEMP. VAL#CC) LitoL,
184
181
1'74
173
182
180
241
236
215
213
188
177
105
.........
240
3 LOWEST
E12
TEMPEATURCES
27MS
gn,
SIT
-.
-4r
I--
-~
~~--
446----
-4
I4
46
'240--
IM
'I
.E
RE'28-
-A
-2
3t-
itit4
'29
?55
ERANGATO, JOHN
-sON,
2V
Is
1 0
,-4
26
55
-2
6--56
(Continued)
6
Data,
A2 (Sh;etgo 26)
Iidu
U0,0 Tle -A2-Copu#
tF.14#A
-5
'T~~~~~
A EAN,
~ ~--6--
'260il
a-
a2--1
260t
209
240-
2so5
24
i302292
985
4261
2110
80
_P5794.-.--
--
36060 H-48U8 firAP41AUX

TE~o VAL*(C W L*.04..
-296
220
235
25
215
i f3'
20)
4
243
23
40.000 kOUR TkM~i~-NDE"

TiEi*. VA*( -0) .LotoL.
219
215
231
21
li
230
2-11i
208
197
240
239
235
g&9.
45
46
47
48
LISTING NUtISER
43
eJoVr.
S3
SYTE
N~~UpMNE
11
44
sI~3~.A
76
fa4
260
88
94,
56
240
260
260
260
200
2
3
280
300
260
280
230
300'
280
300
280
300
220
LOG*'AVf3' LIFE
AT TE~r
TEMP*CHRS#)
1
2
3
14593
2975
1057
9371
5513
3543
8401
3232
1200
8676 ,11323
1735'
2868
995
1209
7507A
910
264
95% L..G.L*ONLY
VAL~JESCIRS-)
OTEIMPS- 1,2&3
1
2
3
12229
3349
683
9186
5587
3472
7974
3047
1142
8076
2992
1124
7058
2382
615
5660
13.00
198
mE-A,* REG.
3976
3511
1222
1174
804
232
255
253
202
244
158
242
244
243
249
211
24i
i85
30.000 i4VUR TEMihINOXX

TEAiP. VAL.C W) L#C#L9
249
247
199
198
237
234'
237
235
243
234
164
111
fP
E- #IND FX
L.C*Lo
245
243
19U
168
232
2211
2-32
230,
239
229
181
178
LOWEST
TEST
TEPRTA4
IN'E
20.000 HJU1R
tEMP*INDtA
TEMP. VAL-C 0)
40.*0U0 *3tiU
TEme. VAL-( C)
L-d;L.
A'>
40Al
A~
106-
~ ~
As
-*'
-e
fn
/V
W1
10iF MARK ON THIS 5H HT
M A RICK ON (Q- FT-tAY
NRL Rk#ORT 8696
Tatble A2"-ComipUter Readout Data, -(Sheet:2-7) (Co6nin-ued)

LIF LG -AWER-
-49
50-
51 :
-2
SYSTEM NW1EIR
'57-
85--
16
-s
1JL0.A
200'
220:
6
480o_
:200
4056405
A2323
638
_~~ET
42-20
-I-180
JES
TFA ERATURES
LlaGw AVG._'L
LFE,
3-
-220 --
153934
560
522
~T~ TST
Th*.CRS-)
-t
sx L-o5oLaLY
VALC{HS:Y
TEM4PS* f1,2 3
MkAt -R1E(3
-LINE
1
9
3-
30o,000 -- owR TEMP.L'40EX

TP.VAL.(C C)_ L C L.
_40o000, -NOURI -TE~e-oNDEX
-696
_66
48 3'
-3098:455
6?
I0I5.
86
-1-79'
173
F 73-
14
1,68 -
1~631-62-
--
158
1v693'
-25668
1'4#24V
13_23!20
-V18,
122-
1Y64-
ZO 0_
29596
2966
27
JEiPeo VAL6C -)- :Lo-C'L
6
-_280
-5589
A194
-40.
8
691-
20,000 HOUR TEMPoINDEA

-TtI4I- VALol 'C)
.C
100
-235-,232-
r13- -227
-~=A
2
2-401
260_0
220
240
260
-9306- -4410
3342
6
148
4
8950
3429
137
16454
8
1403
1452
-205--
__
-97_1 15
118
19
1_9-7_-46
193
_492-
164
76c
~2
-i-VA1.0,Z
14
!,
i*3;11
13
18
00
22q?
200
20
4002
10
200
720
30
15
iu
lbvD4?d*DEA
TVPi
AL~ C)L.CL.
40'i000
3 LWES 1190
TES2
3
TEMERTUES
LOG _AG.L~F1637
~T ES 22i154
TEMPCKRS) 348
'06
92ITEMPS'
27
12
19
0
2M0
1349778
1158
4152
23
590
119G5
4375
123
14441
24
1239
10003 7905
4547
741
1066
1
2
3
5313
232
407,
5552
67
76
1046527
11684
322
220
358
LLiF
4,60
i113
393
88
8
14
182
-173
182
18
VA.
grio-dHO.
77
20.00 0URTEM.UEX
7
...
)
173
30.00 HOU8TEMPINDX
TEAP* f)EX
40#000 HOU(
TEMPI. VAL.'( C) L..C L.
170
167
km-
Iini
_
+64,
10
200
220-
200
20
2-40
LoC.LsONLY
VALES4~.)
1.2A&3
t4ENREG,.
21
~ME
1175
1050
Ig1o
8
186,1718
!3
18
182
187
182
7
11
18?
8
174
7
1,7 A
6
235
--
--
ff p
ERANtCATO, JQW4S-,ON, CAMPBELL AND ,WALKER,

Table A2-Computer Ridout Data, (Sheet 28)(Contiued)
L1STLNrp~~f~
61
62
SY~STEM NUM5Et
79
81--
5-LWEST
430
Tfr2
1
L&G. A V6- LLFF
At f.ST
2
T~i~r.n,~16324
05
- -.
---
L#C.LoOAILY
-VALUESCi~.
vTI~meSo
2006
220
106913
517't-
$101
315
3064
9300
6123
5 342
406d
1041
3,p263
-1513
LINEJ
17'M9
m'EAN4 REG.
V~90
200
220-
I t6'
20,006 HO(JR fEvWi.NDEA

TE,'Ps VAL.( w)
179
-37-8
1..L77 37
30,600 HOURA T&V - NDEA

TCAP#. VAL*( C) L.C.Lo
173
170
373
372
169
36b
169
168
460000 HO~UR
TEie,~ VAL*( C)
**~ENU
T10
1.'. IlUEJA
L.C.Li.
OF DATA
20.37 SEGSs
u*
fIonfl
20o)6
b&
~~AWib~OOOHUR
--
T~
VM..'d)
flU
1-ifW
4,1
TEAlsnoEA
LiCLo
--
7X7
I??.
-is,-
____
11
31
rW
~
NRL REPORT 8095
Table AS-Generic Name Index for Magnet Wire Insulation
Index No.
[Generic
Name
A
Polyvinyl Formal
B-i thru B-7
Polyester
C-i thru C-5
Modified Polyester
DiD2,D3Modified Polyester with Linear Polyester Topcoat
b
E-1, E-2
F-1, P-2
thru G-8
WG-1
.1, D2, D-
Trio-Polyester
Tris-Polyester with Linear Polyester Topcoat
Tris-Polyester with Linear Polyamideimide
Topcoat4
H-i thru H-il
I
J-1, J-2
K-2, K-S
L-1 thru L-5
M-i, M-2
N-i, N-2
CVK-i,
0
P-1, P-2, P-3
Q
x-R
S
T
U
R
5Z-1,
V
W
X
Y-i, Y-2, Y-3
Z-2, Z-3Epx
Polylmide
Amide-Imide
Nylon
Polyester/Nylon
Bondable
Aromatic Polyester Amide-Imide
Aromatic Polyester Amide-Imide with PolyAmideimide Topcoat
Tris.Polyester with Amide-Imide Topcoat
Polyesterimide
Polyester-Amide
Acrylic
Single Polyester Film/Single Polyester-glass/
Glass/Silicone varnish
Single Polyester Film/Glass/Double Polyester-glass
Single Polyester Film/Dble Polyester-glass/
Silicone Varnished
Polyester Film/Polyester-glass Fiber
Polyester Film/Dble Polyester-glass Fiber
Bare (no film)/Single Polyester-glass Fiber
Silicone Modified Polyester
109
'R
___ ---
Table A4 -Generic Name Index for Insulating Varnishes

Generic Name
Index No.
1A
Oleoresinous, Organic
2A thru 2F
Oil Modified Phenolic
3A
Oil Modified Alkyd
4A thru 4F
Modified Polyester
5A thru 5E
Phenolic Modified Polyester
6B, 60
Silicone Modified Polyester
ER6A,
7A
Tris-Polyester
8A
Diphenyl Oxide Polymer
9A
Unmodified Epoxy
10A, 10B
Silicone
11A
Experimental Solventless
12A
Solventless, Two Component Epoxy
13A, 13B, 13C, 13D
Polyimide (129o solids)
14A
Amide.Imide
15A
Acrylic
110

Reliability Prediction Studies On Electrical Insulation Navy Summary Report NAVAL

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Reliability Prediction Studies On Electrical Insulation Navy Summary Report NAVAL

Uploaded by

Copyright:

Available Formats

'T-77

NRL Report 8095

Reliability Prediction Studies on 1

Naval Ship EngineeringCenter

July 13, 1977

NAVAL RESEARCH LABORATORY

SECURITY CLASSIFICATION OF THIS PAGE ("oen Data Entered)

REPORT DOCUMENTATION PAGE

2. GOVT ACCESSION NO.

3. RECIPIENT'S CATALOG NUMBER

NRL Report 8095

PERFORMING ORG. REPORT NUMBER

INSULATION: NAVY SUMMARY REPORT

S. CONTRACT OR GRANT NUMBER(&)

E. L. Brancato, L. M. Johnson, F. J. Campbell, and H. P. Walker

10. PROGRAM ELEMENT. PROJECT. TASK

Naval Research Laboratory

NRL Problem E02-04.801

AREA & WORK UNIT NUMBERS

12. REPORT DATE

I1. CONTROLLING OFFICE NAME AND ADDRESS

July 13, 1977

13. NUMBER OF PAGES

MONITORING AGENCY NAME & ADDRESS(If diltc. .. t Itm Controlling Office)

15. SECURITY CLASS. (of Chfo repot)

Approved for public release; distribution unlimited.

17. DISTRIBUTION STATEMENT (of lhb.,

ABSTRACT (Contiue on reverse side If necessary and Identify by block number)

EDITION OF I NOV 6- IS OBSOLETE

-,tL1JNITY CLASSIFICATION OF THIS PAOE(Whon Data E~ntered)

insulation system motorette tests are documented.

ORIGIN OF THERMAL CLASSIFICATION AND

PHILOSOPHY OF THERMAL CLASSIFICATION .........

THE NEED FOR THERMAL EVALUATION

DEVELOPMENT OF EXPERIMENTAL EQUIPMENT ......

Magnet Wire Twist Test Procedure ..................

TREATMENT OF DATA .............................

Failure Criteria .................................

CONCLUSIONS AND RECOMMENDATIONS .............

Appendix A-Use of Computer Readout Tables Al and A2 ...

RELIABILITY PREDICTION STUDIES ON ELECTRICAL

Manuscript submitted December 9,1976.

BRANCATO, JOHNSON, CAMPBELL AND WALKER

life-vs-temperature characteristic curve. A detailed statistical- analysis validated truncated

ORIGIN OF THERMAL CLASSIFICATION AND EVALUATION

need to know the effects of temperature on electrical insulation was first

NRL REPORT 8095

Fig. 1-Possible life-vs-temperature relationship of

BRANCATO, JOHNSON, CAMPBELL AND WALKER

tion of insulation failure by current flow, and (c) measurement of temperature by

Fig. 2-Coil assembly showing asbestos wrapping and lead attachment

NRL REPORT 8095

BRANCATO, JOHNSON, CAMPBELL AND WALKER

Formex insulated coils at 1600 C,

95% confidence intervalA

Fig. 4-Effect of vibration on life of

NRL REPORT 8095

Fig. 5-Assembled motorette specimen with

PHILOSOPHY OF THERMAL CLASSIFICATION

BRANCATO, JOHNSON, CAMPBELL AND WALKER

with phenolic varnishes.

IIVES AND 95% LOWER ONLY CONFI-

130 140 150