TECHNICAL

NASA

NASA TN D-5390

NOTE

o_
|

Z
I.,-,-

EFFECT
ON
IN

OF STRESS

RATIO

FATIGUE-CRACK
7075-T6

AND

2024-T3

ALUMINUM-ALLOY
by C. Michael

GROWTH
SPECIMENS

Hudson

Langley

Research Center

Langley

Station,

Hampton,

Va.

NATIONAL AERONAUTICS AND SPACE ADMINISTRATION

WASHINGTON, D. C.

AUGUST 1969

,1

1.

Report

No.

2.

Government

Accession

No.

3.

Recipient's

S.

Report

Catalog

No.

NASA TN D-5390
4.

Title

and

Subtitle

AND 2024-T3ALUMINUM-ALLOY SPECIMENS

7.

Author(s)

6.

Performing

Organization

Code

8.

Performing

Organization

Report

C. Michael Hudson
9.

Performing

Date

August 1969

EFFECTOF STRESSRATIO ON FATIGUE-CRACKGROWTHIN 7075-T6

No.

L-6662

Organization

Name

and

10.

Address

Work

Unit

No.

126-14-15-01-23

NASA Langley Research Center

111.

Contract

or Grant

No.

Hampton, Va. 23365

13.
Sponsoring

Agency

Name

and

Type

of

Report

and

Period

Covered

Technical Note

Address

National Aeronautics and SpaceAdministration

Washington, D.C. 20546
14.

15.

Supplementary

16.

Abstract

Sponsoring

Agency

Code

Notes

Axial-load fatigue-crack-growthtests were conducted on 12-inch-wide (305-ram)sheet specimensmade

of 7075-T6 and 2024-T3aluminum alloy. These tests were madeat stress ratios R (ratio of the minimum
stress to the maximum stress) ranging from -1.0 to 0.8 and at maximum stress levels ranging from 5 to
50 ksi (34 to 340 MN/m21 to study the effectsof stress ratio on fatigue-crack growth.
The experimental results were analyzed by using the stress-intensify

method. For a given positive

stress ratio, the fatigue-crack-growth rate was a single-valued function of the stress-intensity range for
both 7075-T6 and 2024-T3aluminum alloys. For R >- 0 the crack-growth rates varied systematically with R
for both materials; the higher stress ratios produced higher rates of fatigue-crack growth for a given stressintensity range.
Fatigue cracks in the 7075-T6 aluminum alloy grew at the same rates in all tests with R -< 0 when
the same maximum stress-intensity

factor was applied. In contrast, fatigue cracks in the 2024-T3 alumi-

num alloy grew faster in the tests with R = -1 than in the tests with R = 0 when the same maximum
stress-intensity factor was applied.
Empirical equations previously developedby various investigators were fitted to the experimental data.
In general, good correlation was obtained.

17.

Key

lB.

WordsSuggested by Author(s)

Distribution

Statement

Fatigue
Unclassified- Unlimited

Crack propagation
Stress ratio

19.

Security

Classif.

Unclassified

(of

this

report)

20,

Security

Classif.

Unclassified

(of

this

page)

21.

No.

of

Pages

29

_For sale by the Clearinghouse for Federal Scientific and Technical Information
Springfield, Virginia 22151

22.

EFFECT

OF STRESS

7075-T6

RATIO

AND 2024-T3

ON FATIGUE-CRACK

GROWTH

ALUMINUM-ALLOY

By C. Michael
Langley

IN

SPECIMENS

Hudson

Research

Center

SUMMARY

sheet

Axial-load

fatigue-crack-growth

specimens

made

at stress
-1.0

to 0.8 and at maximum

stress

effects

The
a given
of the
the

and

of the

the

of 7075-T6
(ratio

study

ratios

tests

of stress

experimental

positive

stress

crack-growth
ratios

produced

minimum

aluminum

stress

levels

were

ratio,

the

range

for both

rates

2024-T3

conducted

ranging

on fatigue-crack

results

stress-intensity

stress

ratio

were

alloy.

rates

were

made

stress)

from

(34 to 340 MN/m2)

5 to 50 ksi

by using

7075-T6

higher

tests

to the maximum

ranging

the stress-intensity

fatigue-crack-growth

systematically

These

(305-mm)

from
to

growth.

analyzed

varied

on 12-inch-wide

rate

and

was

2024-T3

with

growth

aluminum

grew

a single-valued

aluminum

for both

of fatigue-crack

method.

for

function

alloys.

materials;

For

For

R -> 0

the higher

a given

stress-intensity

range.
Fatigue
with

cracks

R -<-0

fatigue

when

cracks

in the

tests

vol.

entitled
fitted

with

2024-T3

data

except

stress-intensity
alloy

the

maximum

same

developed
1967),

by Erdogan

Forman's

data.

Erdogan's

at the higher

and

growth

faster

et al.

Paris'
rates

in the

the

tests

7075-T6

In contrast,

factor

and

good

R = -1
was

ASME,

Ser.

by Paris
Press,

fit to both

showed

in all tests

with

Univ.

an excellent

rates

applied.

(in Trans.

Syracuse

equations
for

was

(in NASA CR-901),

produced

same

stress-intensity

Approach,"

equation

at the

factor

grew

by Forman

An Interdisciplinary

data.

alloy

aluminum

when

89, no. 3, Sept.

to the

maximum

2024-T3

R = 0

7075-T6

same

equations

"Fatigue

the

the

in the

Empirical
Eng.,

in the

the

correlation

than

applied.
D:

J. Basic

(in book
1964)

were

7075-T6
with

and
the test

alloy.

INTRODUCTION

Fatigue
components.
spent
shown

cracks

frequently

Consequently,

in the crack-propagation
to be dependent

upon

initia_te
a major
phase

the applied

early

portion

in the

life

of the useful

of fatigue.

of cyclically
life

Fatigue-crack

stress-intensity

range

loaded

of these

components

propagation
and

upon

structural

the

has
stress

is
been

ratio R (ratio of the minimum stress to the maximum stress). While much information has been obtainedfor various stress-intensity ranges, much less information is
available with regard to stress ratio. Accordingly, an investigation has been conducted
to determine the effects of a wide range of R values and stresses on fatigue-crack
growth in 7075-T6 and 2024-T3 aluminum-alloy sheet specimens. These materials were
selected becauseof their frequent use in aircraft construction.
The data were analyzedby using the stress-intensity method. Figge and Newman
(ref. 1) showedthat by this methodthe data from simple sheet specimens could be used
to predict fatigue-crack-growth behavior in simulated structural configurations. Empirical equationsdevelopedby Forman, Kearney, and Engle (ref. 2), by Erdogan (ref. 3), and
by Paris (ref. 4) were fitted to the data generatedin this investigation by using leastsquares techniques.
SYMBOLS
The units usedfor the physical quantities defined in this paper are given both in
U.S. Customary Units and in the International Systemof Units (SI) (ref. 5). The appendix
presents factors relating these two systems of units.
one-half of total length of a central symmetrical crack, inches (mm)
af

half-length of crack immediately prior to rapid fracture, inches (mm)

ai

half-length of crack at onset of slow crack growth, inches (mm)

C,C1,C2

constants

kc

critical

kmax

stress-intensity

in fatigue-crack-growth

stress-intensity

kmin

factor

factor

tangent-formula
stress-intensity

equations

width
factor

at failure,

corresponding

range

of stress-intensity

correction),

ksi-inl/2

corresponding

kmax-

m,n,p

exponents

kmin,

factor

ksi-inl/2

in fatigue-crack-growth

(MN/m3/2)

to maximum

cyclic

stress

(with

stress

(with

(MN/m3/2)

to mininmm

tangent-formulawidthcorrection),ksi-inl/2
Z_k

psi-inl/2

cyclic
(MN/m3/2)

(with

ta-_gent-formula

_MN/m3/2)
equations

ijr

width

correction),

number

of cycles

Pa

amplitude

Pf

load

on specimen

immediately

Pi

load

on specimen

at onset

Pm

mean

Pmax

maximum

load

Pmin

minimum

load

ratio

Sa

alternating

Sm

mean

Sma x

maximum

gross

stress,

Pmax/Wt,

Smin

minimum

gross

stress,

Pmin/wt,

specimen

thickness,

specimen

width,

length

O/

correction

of load

load

applied

applied

applied

Pm/wt,

ksi

inches

starter

growth,

kips

kips

kips

(newtons)

(newtons)

(newtons)

Pm - Pa,

ksi

fracture,

(newtons)

to maximum

kips

(newtons)

stress

(MN/m2)

(MN/m2)

inches

for finite

crack

kips

Pa/wt,

(newtons)

to rapid

in a cycle,

stress

stress,

of crack

of slow

in a cycle,

applied

kips

prior

in a cycle,

of minimum

stress,

in a cycle,

ksi

ksi

(MN/m2)

(MN/m2)

(mm)

(mm)

notch,

width

SPECIMENS,

inches

(mm)

of panel

TESTS,

AND PROCEDURES

Specimens
The

materials

aluminum-alloy
fatigue
were

properties
obtained

were

sheets

retained

of these
in this

taken

from

at Langley

materials

investigation

the special

are
by using

stocks

Research
discussed
standard

of 7075-T6

Center

for

and

fatigue

in reference

6.

American

Society

2024-T3
testing.

Tensile
for

The

properties
Testing

and
3

Materials (ASTM) tensile specimens andare listed in table I. Also listed in table I are
the tensile properties obtainedin 1949on the same stocks of material. The tensile
properties of the materials did not changesignificantly over the 19-year interval. The
specimen configuration used in the crack propagation andin ancillary residual-staticstrength tests is shownin figure 1. Sheetspecimens 12inches (305mm) wide, 35 inches
(889 mm) long, and with a nominal thickness of 0.090inch (2.28 ram) were tested. A
notch 0.10 inch (2.54mm) long by 0.01 inch (0.25 mm) wide was cut into the center of
each specimenby an electrical discharge process. Only very localized heating occurs
in making notchesin this manner. Thus virtually all of the material through which the
fatigue crack propagatesis unaltered by the cutting process. All specimenswere made
with the longitudinal axis of the specimensparallel to the rolling direction of the sheets.
A reference grid (ref. 7) was photographically printed on the surface of the specimens to mark intervals in the path of the crack. Metallographic examination andtensile
tests conductedon specimens bearing the grid indicated that the grid had no detrimental
effect on the material.
Testing Machines
Four axial-lo_id fatigue-testing machineswere employedin this investigation. The
capabilities of these machines are listed in the following table:
Maximum load
capability
lbf
kN

Machine type

cpm

Machine
description

Hz

Subresonant ........
Hydraulic .........
Hydraulic .........

20 000
100000
120000

89
445
534

1800
1200
30

30

Reference

20

Reference

Reference

10

Combination:
As subresonantunit .
As hydraulic unit ....

105000
132000

467
587

820

14
Reference

11

Loads

were

continuously

monitored

strain-gage

bridge

attached

to a dynamometer

mum

in loading

error

was

1 percent

Axial-load
ranging
aluminum

from

fatigue-crack-propagation
-1.0

alloy.

to 0.8 for
Generally,

_075-T6
tests

on these

0.7 to 1.0

machines

in series

by measuring

with

the

the

specimens.

output
The

load.

aluminum
conducted

were
alloy

conducted
and

from

at a number

at stress
-1.0

ratios

to 0.7 for

of maximum

2024-T3

stress

of a

maxi-

Procedure
tests

were

0.5

40 to 60

of the applied
Test

Operating frequency

levels Smax for a given stress ratio. The alternating and meanloads were kept constant throughout eachtest. Duplicate tests were conductedat each stress level.
Fatigue-crack growth was observed through 10-power microscopes while illuminating the specimenwith stroboscopic light. The number of cycles required to propagate
the crack to eachgrid line was recorded so that crack-propagation rates could be determined. Approximately two-thirds of the crack-propagation tests were conductedto failure. The remaining one-third were stoppedbefore failure, andthe cracked specimens
were used in residual-static-strength tests.
In all the tests (crack growth and residual static strength), the specimenswere
clampedbetweenlubricated guidesin order to prevent buckling and out-of-plane vibrations during testing. Light oil was used to lubricate the surfaces of the specimensand
the guides. Noneof this oil was observedto enter the crack during testing. Consequently, the oil was not expectedto affect the crack growth. A 0.125-inch (3.2-mm) slot
was cut across the width of one guide plate to allow visual observation of the crack.
Axial-load residual-static-strength tests were conductedat a load rate of
120000lbf/min (8.9 kN/s) on unfailed crack-propagation specimens. A 70-mm
sequencecamera operating at 20 frames per secondwas used to obtain slow-crackgrowth data. The cracked section of the specimen andthe image of a load-indicating
device were photographedon eachframe of film by using an optical prism. From this
film, the load at which the crack first started to grow statically and the load and crack
length immediately prior to final failure were determined.
METHODOF ANALYSIS
The fatigue-crack-growth datawere correlated by the stress-intensity method. It
was hypothesizedin reference 4 that the rate of fatigue-crack propagationwas a function
of the stress-intensity range; that is,
d_._.a
= f(_k)
dN

(1)

Ak = kma x - kmin

(2)

where

For

centrally

cracked

specimens

subjected

kmax

to a uniformly

= aSmax_Ja

distributed

axial

load,

(3)

and
kmin

The term
by

is a factor

O/

which

= (_Smin_-a

corrects

for

(4)

the finite

width

of the

specimen

and

is given

a = i_- _ tan 7raw

The

term

Sma x

gross

stress

were

plotted

is the maximum

in the

cycle.

as functions

gross

stress

In presenting
of

Ak

in the

the results,

(eq.

are

presented

crack

to grow

from

II.

This

table

a half-length

Fatigue-crack-propagation
growth

curves

defined

Typical
under

in table

two to four

loading

times

alloy

as in the

gives

presented
For

are

cycles

for

materials;
a given
Data

for

of

loading

cycle

equation
as the

of

for

was

for

of cycles
to the

specimens

required

specified

determined

for

the

half-lengths.

from

the

crack-

7075-T6

and

in figure

required

2.

to reach

2024-T3

For

specimens

these

a given

tested

identical

crack

conditions,

length

in the

alloy.

ratio,

of Stress

Ratio
data

as a function

rate

was

stress

ratios

from

of the

rates
produced

tests

function

varied
higher

the

with

R = 0

stress-intensity

a single-valued

Crack-growth

The

crack-growth

all negative

neglected

(2) became
data

da/dN

of

range
of

systematically
rates

Ak
with

of fatigue-crack

are
Ak.

for both
R

for

growth

Ak.

R =<0.-

Ak

ram)

graphically

for

shown

were

of rate

alloys.

the higher

value

function

stress

2024-T3

minimum

values

and 2024-T3

number

(2.54

fatigue-crack-growth

3 as plots

positive

and

The

inch

were

curves

7075-T6

R ->-0.-

in figure

a given

7075-T6
both

for

experimental

on 7075-T6

the average

of 0.10

Effect
Data

is the

II.

conditions

as many

tests

da/dN

fatigue-crack-growth

identical

2024-T3

rates

Smin

Growth

of the fatigue-crack-growth

in table

the

and

AND DISCUSSION

Fatigue-Crack
results

cycle

(2)).

RESULTS

The

(5)

R = 0

kmax.
(for

stress

rate
ratios

in calculating
These
which

data
hk

when
Ak

for
also

in the

(fig.

negative
equals

7075-T6
the

alloy

was

compressive

4).

That

R
kmax),

fell

portion

is, for
into

a single-valued

R < 0,

the same

indicating

of the
Ak
scatter

the compressive

in
band

portion of the loading cycle did not significantly affect fatigue-crack growth in 7075-T6
alloy.
The crack-growth rate in the 2024-T3 alloy was nominally a single-valued function of Ak for the negative stress ratios (e.g., see fig. 5). (Rates do appear to be
slightly higher in the low-frequency tests than in the high-frequency ones for this set of
data.) However, fatigue cracks in the 2024-T3 alloy grew faster in the tests with R = -1
than

in the tests

and

R = -1

intensity
crack

with

data.

range

and

closed

completely

under

compressive

the

at

would

and McEvily
pletely

when

tion at the

zero

tend

to close

rial

immediately

higher

load
tip.

the

Thus,

is reached
Thus,

(ref.

and,

equations.

equations
R < 0
data
were
tive

are
tests

for

for

R =<0

not used

compressive

(fig.

were

the

R=

tensile

cycle

stress-

accelerated

alloy

This

that

the equations

since

a crack

(ref.

there

Illg
com-

deforma-

R = -1
damage

could

and

would

to the

mate-

be manifest

as the

Equations
by Forman,

4) were

fitted

the appropriate
given
these

to apply

which

in the

only.

of plastic

with

fatigue

developed

band.

cycle

investigation.

In fitting

assumed

existed

do not close

because

damage

the constants

scatter

specalloy

no crack

loading

alloy

Fatigue-Crack-Growth

to determine

7075-T6

loading

R = 0

in the tests

additional

in this

Units.)
were

an

R = -1

2024-T3

at the tip of a crack,

do additional

3), and by Paris

same

for

with

equations

used

of the

in 2024-T3

loading

and

as though

portion

cracks

so, would

With

be noted

into the

factor

same

in the

to the material

tests

tip.

cracks

experiencing

observed

Customary

in fitting

both

was

is open

data

constants

in this

all

R = 0

data.

for

the

tests

these
these

data

from

since

alloy

all

with

of calculating

an indefinable

and

for

report

for 2024-T3

no method
for

to the test

equations,

to the

The

Kearney,

the

the
R = -1

the effec-

portion

of the

loading.

Forman's

data

fell

stress-intensity

intensity

(ref.

7075-T6

the

7075-T6

behaved

compressive

that

in doing

of Data

(It should
for U.S.

for

loading

for

that

material

compressive

rates

techniques

for

of the

proposed

it were

of the crack

2), by Erdogan

tests

findings

no damage

fatigue-crack-growth

Least-squares
three

the

in the

the

Correlation

Engle

the

observed

fatigue-crack-growth

Empirical

and that

as though

ahead

portion

They

do virtually

cracks

Ak=kma

R = -1

similar

R = -1.

12) further

crack

again,

in the

compression

load

propagate

(ref.

Here

12) reported

loading.
would

6).

alloy.

and

at zero

fig.
rates

the

(ref.

R = 0

R < 0

crack

higher
that

McEvily

tested

with

(see

in the 2024-T3

imens

tests

These

indicate

growth
Illg

R=

factor
7).

equation
at failure),
This

equation

(ref.

2), relating

produced
has

da/dN,

an excellent

the form

Ak,

R, and

fit to both

the

kc

(the

7075-T6

critical
and

the

stress2024-T3

da _
C(Ak)n
dN (1 - R)kc - Lik

(6)

where C and n are empirically determined constants. The values of kc for

7075-T6 and 2024-T3 alloys were obtainedfrom the auxiliary residual-static-strength
tests. This factor is related to the load at rapid fracture Pf, the associated crack
length af, and the width-correction factor ot, as follows:

Average

values

of

56 600 psi-in1/2
(table III).
The
lowing

kc

of 40 400 psi-inll

MN/m3/2)

2 (44.4

(62.2 MN/m3/2)
for the 2024-T3
constants
C and
n in equation

the

7075-T6

alloy

and

of

alloy were found in these

tests
(6) were determined
to have the fol-

values:
Material

The
fig.

for

8).

equation
This

.....

2.13 10 -13

3.21

2024-T3

.....

3.22

3.38

has

by Erdogan
the

(ref.

C1,

the

data,

test

m, and
these

are

constants

10 -14

3) was

also

fitted

to the test

data

(see

form
d_._a=
m
dN
C lkmax

where

7075-T6

developed

equation

empirically
were

determined

determined

(8)

Akp

constants.

to have

In fitting

the following

equation

(8) to

values:

C1
1.00 10 "19

7075-T6
.....
Material
2024-T3
.....

The

curves

correlation
Paris
growth

computed
with

and

(ref.
the

the

by equation
test

data

4) proposed
stress-intensity

1.04 10 -19

(8) are

except

at the

the following
range

shown

in figure

higher

growth

relationship

(in the

notation

da = C2(Zik)4

1.15

8.

This

rates
between

of the

2.44
equation

for the
the rate

present

showed

7075-T6

good

alloy.

of fatigue-crack

paper):

(9)

In equation (9) C2 is a constant which is proposedto incorporate the effects of material, meanload, loading frequency, and environment. This equationalso showedgood
correlation with the test data except at the higher growth rates for the 7075-T6 alloy (see
fig. 9). Separatevalues of the coefficient C2 had to be computedfor eachvalue of R
since R is not a function in equation (9). These coefficients are listed in the following
table:
Material

C2

7075-T6 ......

5.52 10-21
6.44
1.00 10-20
1.80
3.95
6.84

.2
.33
.5
.7
.8
0

2024-T3 ......

2.14 10-21
5.40

.33
.5

q.75

.7
The
ture.

7075-T6

A reflex

approaching
primary
equation.
had

tests

in figures

curvature

is also

(1 - R)kc

excellent

The

data

for

the 2024-T3

been

conducted
(1 - R)kc).

well

being

equations

equation

as Forman's

intensity

Forman's

7075-T6
would

high

or reflex

equation;
(6).

data

This

obtained

probably

have

investigation

of curva-

it is induced

by

intrinsic

shape

by using

Forman's

had

stress-intensity

type

a reflex

ranges

of crack-growth

Ak

is the

curvature

(such

that

behavior

Z_k

at very

high

conducted.

and Paris'

these

alloy

A separate

an "S" shape

of equation

fit to the

at sufficiently

into

from

in the denominator

the

Erdogan's

Forman's

obtained

for

is currently

quently,

7 to 9 fell

reason

approached
rates

data

1.24 10 -20

equations
cannot

did (see
equation

do not provide

fit the

figs.

7075-T6

7 to 9) and

if there

were

for this

data

reflex

at the high

probably

additional

would
data

curvature.

growth

not fit the

from

tests

Conse-

rates

as well

as

data

as

2024-T3

at high

stress-

ranges.
CONC LUSIONS

Axial-load
12 inches
2024-T3
stress

fatigue-crack-propagation

(305 mm)
aluminum

to the

wide
alloys.

maximum

and nominally
These

stress)

tests

ranging

tests
0.090

were

inch

conducted

(2.28

were

at stress

from

-1.0

mm)
ratios

to 0.8 and

on sheet
thick
R

made
(ratio

at maximum

specimens
of 7075-T6
of the

and

minimum

stresses
9

ranging

from

growth.

The

correlated
this

5 to 50 ksi
test

with

results

three

were

analyzed

empirical

to study

by using

relations.

The

the

the

effect

of

stress-intensity

following

fatigue-crack

on

method

conclusions

can

and were

be drawn

from

study:
1. For

intensity

a given

range
2. For

given

positive

for both
R ----0

materials.

The

the
the

in this

and

stress

rate

2024-T3

was

function

of stress-

alloys.

rates

ratios

a single-valued

varied

produced

systematically

higher

rates

with

of fatigue-crack

for

both

growth

for a

range.
cracks

same

ratio,

fatigue-crack-growth

stress-intensity

when

stress

7075-T6

higher

3. Fatigue

that

(34 to 340 MN/mZ)

in 7075-T6

maximum

compressive

alloy

grew

stress-intensity

portion

of the

at the

factor

loading

was

cycle

same

rates

applied.

did not

in all
These

significantly

tests

with

R <=0

equal

rates

indicate

affect

crack

growth

R = -1

than

material.
4. The

fatigue

cracks

in the tests

with

R = 0

Apparently,

the

compression

in 2024-T3
when

the

alloy

same

portion

grew

faster

maximum

in the tests

stress-intensity

of the loading

cycle

with
factor

accelerated

was

crack

applied.

growth

in this

material.
5. For
to reach

identical

a given

Eng.,

(in book

entitled

were

fitted

conditions,

two to four

length

in 2024-T3

alloy

crack

6. Empirical
J. Basic

loading

equations

vol.

developed

89, no. 3, Sept.

"Fatigue

to the data.

and

the 2024-T3

data.

test

data

at the higher

Langley

except
Research

National

Erdogan's

equation
and

growth

Station,

and

Space

Hampton,

126-14-15-01-23.

10

by Erdogan

rates

produced

the

May

Syracuse

showed

7075-T6

Administration,
Va.,

(in Trans.

an excellent

equations
for

were

ASME,

Set.

29, 1969,

required

alloy.

Approach,"

Paris'

cycles

(in NASA CR-901),

Center,

Aeronautics
Langley

et al.

An Interdisciplinary

Forman's

as many

as in 7075-T6

by Forman

1967),

times

and by Paris
Univ.

fit to both
good

alloy.

D:

correlation

Press,
the

1964)

7075-T6
with

the

APPENDIX
CONVERSIONOF U.S. CUSTOMARYUNITS TO SI UNITS
The International Systemof Units (SI) was adoptedby the Eleventh General Conference of Weights and Measures, Paris, October 1960,in Resolution No. 12 (ref. 5).
Conversion factors for the units usedherein are given in the following table:
To convert from
U.S. Customary Units

Multiply by -

lbf
in.
ksi
cpm
Prefixes

and

symbols

4.448222

newtons

2.54 10-2

meters

6.894757 x 106

newtons/meter2

1.67 x 10-2

hertz

to indicate
Multiple
10-9

To obtain SI units

multiples
Prefix

of units

are

(N)
(m)
(N/m2)

(Hz)

as follows:

Symbol

nano

10-3

milli

103

kilo

106

mega

109

giga

11

REFERENCES
1. Figge, I. E.; andNewman,J. C., Jr.: Fatigue Crack Propagation in Structures With
Simulated Rivet Forces. Spe.Tech. Publ. No. 415, Amer. Soc. Testing Mater.,
c.1967, pp. 71-94.
2. Forman, R. G.; Kearney, V. E.; andEngle, R. M.: Numerical Analysis of Crack
Propagation in Cyclic-Loaded Structures. Trans. ASME, Ser. D: J. Basic Eng.,
vol. 89, no. 3, Sept. 1967,pp. 459-464.
3. Erdogan,
4. Paris,

Fazil:
Paul

Crack

C.:

The

disciplinary

Univ.

Com.,

6. Grover,

Mar.

H. J.;

7. Hudson,

Aluminum

Alloys

8. Grover,

H. J.;

9. McEvily,
Two
10. Illg,

J.,

Walter:
7075-T6

the

Life

11. Hudson,

Jr.;

and Illg,
NACA

for Two
1959.

12

An Inter-

Weiss,

Tests

Rep.

eds.,

102,

U.S.

1958.

From

2 to 10,000

Cycles.

Hardrath,

Herbert

and

Loadings

on Fatigue

Specimens
Special

TN

3866,

1956.

Propagation

TN 2928.)

Propagation

With

of the

F. M. :

as Determined

NACA

Sheet

Investigation

Crack

Howell,

Alloy

Steel

NACA
F.:

and Stainless-

of Fatigue-Crack

and Unnotched

and of SAE 4130

B.; and

(Supersedes

Rate

1951.

Titanium

Aluminum

1954.

The

Aluminum

Alloys

Arthur
Under

J.,

Jr.:

The

Completely

Rate

Effects

Patterns.

of Fatigue-Crack

Reversed

Loading.

in

of 2024-T3
Consideration

of
NASA

1963.
and McEvily,

and

1964.

75S-T

Walter:

of 24S-T3

TN 2324,

Charles

and

of Aircraft

Specimens

in Several

TN 4394,

on Notched

Sheet

TN D-2331,

1190,

Strengths

NACA

Landers,

Alloys

C. Michael;

Walter;

NACA

Alloys.

Variable-Amplitude

12. Illg,

Paul;
of 24S-T

Aluminum

Range

TN D-1803,

NASA

Properties

Fatigue

and

on Unnotched

Propagation

W. S.; Kuhn,

Laboratories.

Arthur

NBS Handbook

Fatigue

and of SAE 4130 Steel.

One Superalloy.

Hyler,

Aluminum

Fatigue

and Volker

Guide.

L.R.:

Tests

Fatigue-Crack

Fatigue

in Several

to Fatigue.
L. Reed,

Practice

Jackson,

Fatigue

Alloys

and

Axial-Load

Metric

S. M.; and

C. Michael:

Steel

1967.

pp. 107-132.

ASTM

Axial-Load

75S-T6

Norman

CR-901,

10, 1967.

Bishop,

Materials.

NASA

Approach

J. Burke,

1964,

Pract.:

Theories.

Mechanics

John

Press,

on Metric

Dep.

Fracture

Approach,

Syracuse
5. Comm.

Propagation

Propagation
NASA

TN D-52,

of

_'_

o_O ._

._

_._

:_

_'

_.1.7 _

O0

_1_

0
D_

_D

r_

1._

_'_

t._

0
L_-

13

Z
o

_o_oo

ooooo

oo

14

oo_oo

L_ooo

_o_oo

o_oooo

_o

_oo

oo

,.9,o

_ooooo

ooooo

oo

o
o.
o

o
Z
o

<

r_
Z

0
t_

v
0
p.l

ii

co
o

_g

o
11

ii

15

TABLE

HI.-

CRACK

AND

LOAD

MEASUREMENTS

AND VALUES

RESIDUAL-STATIC-STRENGTH

(a)

af

ai
in.

7075-T6

mm

FROM

alloy

pf

kc

mm

kips

kN

kips

kN

ksi-inl/2

MN/m3/2

1.72
1.59
1.53
1.94
1.02
.98
.91
1.01
1.05

43.69
40.39
38.86
49.28
25.91
24.89
23.11
25.65
26.67

1.85
1.80
1.79
2.36
1.16
1.19
.96
1.09
1.15

46.99
45.72
45.43
59.95
29.46
30.22
24.38
27.69
29.21

23.4
17.7
20.0
16.5
....
....
....
....
29.2

104
79
89
73
--------130

29.9
30.4
30.4
27.2
36.5
38.8
43.1
41.7
37.7

133
135
135
121
171
173
192
185
168

39.2
39.2
38.5
39.8
38.0
39.7
38.0
41.4
38.4

43.1
43.1
42.3
43.7
41.7
43.7
41.7
45.5
42.2

.78
1.31
1.90
.52
1.22
1.76
1.40
1.17
.54

19.81
33.27
48.26
13.21
30.99
44.70
35.56
29.71
13.72

.94
1.75
2.19
.73
1.53
2.10
1.73
1.43
.71

23.88
44.45
55.63
18.54
38.86
53.34
43.94
36.32
18.03

32.8
19.2
15.7
32.8
25.8
18.3
26.8
29.2
33.3

146
85
70
146
115
81
119
130
148

44.5
32.0
25.0
50.4
35.9
30.2
32.3
36.4
49.8

198
142
111
224
160
134
144
162
222

41.6
40.2
35.4
40.4
42.0
41.8
41.0
40.2
38.5

45.7
44.2
39.0
44.4
46.2
45.9
45.0
44.2
42.3

1.51
2.25
2.47
1.78
.70
.53
1.07
1.12
.55

38.35
57.15
62.74
45.21
17.78
13.46
27.18
28.44
13.97

1.65
2.73
2.81
2.13
.93
.78
1.20
1.37
.72

41.91
69.34
71.37
54.10
23.62
19.81
30.48
34.80
18.29

24.0
12.3
14.0
14.8
....
29.4
30.O
3O.O
....

107
54
62
66
--131
133
133
---

39.0
24.3
22.8
27.1
47.1
48.4
39.0
36.0
47.5

173
108
101
121
210
215
173
160
211

47.6
41.0
39.0
39.0
43.4
40.3
48.2
39.4
37.6

52.3
45.0
42.9
42.9
47.7
44.3
52.9
43.3
41.3

(b) 2024-T3

ai

af

ln.

mm

50.3
68.6
65.5
62.2
61.7
39.6
46.5
57.2

2.82
3.24
3.28
3.3O
3.06
2.42
2.72
2.88

I
i
I
I
I
I
I
I

71.6
82.3
83.3
83.8
77.7
61.5
69.1
73.2

56.6
39.9
45.7
38.6
51.3
40.1
65.0
38.6
38.1

3.24
2.32
2.58
2.38
2.48
2.28
3.22
2.32
2.38

I
I
]
]
]
I
I
I
I

82.3
58.9
65.5
60.5
63.O
57.9
81.8
58.9
60.5

mm

1.98
2.70
2.58
2.45
2.43
1.56
1.83
2.25

I
I
I
I
I
I
I
]

2.23
1.57
1.80
1.52
2.02
1.58
2.56
1.52
1.50

1
]
1
1
I
i
I
t
1

not be

determined

from

alloy

(a)

film.

kc

Pf

Pi

In.

aCould

16

kc

TESTS

Pi

in.

OF

kips

kN

ksi-inl/2

32.0
26.8
28.2
29.5
30.0
36.8
34.7
31.7

142
119
125
131
133
164
154
141

55.9
52.8
55.8
56.9
55.8
58.0
59.0
56.3

61.4
58.0
61.3
62.5
61.3
63.7
64.8
61.9

29.8
36.8
34.2
37.5
34.4
39.7
27.7
38.0
37.0

133
164
152
167
153
177
123
169
165

58.7
54.8
55.5
58.7
54.9
57.5
54.1
57.8
59.0

64.5
60.2
61.0
64.5
60.3
63.2
59.5
63.5
64.8

MN/m3/2

6
(153)

-2 _,- 2H
Is1,: (511
/

it
I-i12 (38)
'9
2-I/4 (57)

_L__
_4

(32)

35
(889}

Stress
raiser

I'-

0.10 -_

._i_/(2.s_ /
._rv------ _
0.010
(0.25)

Detail of stress raiser

+
12
{305)

Figure

1.-

Specimen

configuration,

(All

dimensions

are given

first in inches

and parenthetically

in ram,)

1'/

_-E

-r
0

O
N

0
i

:E

Z
i

--

0,,I
o

r-o

8
g_

_0

ii

ii

E
No

m
E
E

%
r_

d,
"o
Z
0
eJ

Z
0

0
I"--

CO
r_

0
"x

"

O
z

E
E

No

q_

ed

c_

c_

c_

d
J_

0-"!-

18

Ak,

.0
I0

MN/m

I0

3/2

20

40

30
I

-I0
0-

jd z
0

-I0

oo
Rote,
I0

-3

nm/cy

--

Rote,

-I0

in./cycle

cle

0
O

I0

Sym bol

-I0

0.2
0.33

id5 -

0.5
0.7

0.8

-6

I0

, t

I0

15

20

L_k, ksi-in

25

I0

30

35

1/2

(a) 7075-T6 alloy.

Figure 3.-

Variation

of fatigue-crack-growth

rate with

Ak for

R >_-0.

19

Ak,MN/m
I0

.z 0

3/2
3O

20

I0

4O

-I0

id s

O
O
4

I0

Rote,
nm/cycle

,d
in./cycle

165
R

Symbol

0.33
0

io-+

0.5
0.7

0
0

I0 0

I0

15

20

Ak,ksi-in

25
I/2

(b) _24-T3 alloy.

Figure 3.- Concruded.

20

3O

35

A kj MN/m3/2

10-1

I0

20

3O

4O

6
-I0
0

0
o o

10-2
0

o
000

oo
5
0

-I0

10-3

Rate,
i n./cycle

-I0

Rate,
nm/cycle

10-4

I0

-5
10
2
I0

10-6

@
o

10

15

20

25

30

35

Ak, ksi-in 1/2

Figure

4.-

Variation

of fatigue-crack-growth

rate with

Ak

for

R < 0 in 7075-T6 alloy.

21

L_k,MN/m

3/2
!

40

3O

20

I0

-I0

i0

-3
II
4

I0
II

t I

Rate,

Ik

-4

nm/cycle

I0

Rote,

I0

la
Q

in./cycle

id 5

Loading

Symbol

frequency

30cpm
820

(0.5 Hz)

cpm(13.7

Hz)
2

-I0

15sI0

I (_7

5I

I0!

i
15

|
20

_k,ksi-in
Figure 5.-

22

Variation

of fatigue-crack-growth

rate with

6k

for

R = -1

25i

30I

35

i/2
in 2024-T3 alloy.

(Data are for two loading frequencies.)

Llk, MN/m
I0

3/2

20

30

40

-I0
(3
t_
m

ld3

QD
Q
r_

0
4

I0

r_O

aa
jO "4

t3

Rate,

s_ g

n m/cycle

_ o%
a

,d

Rate,
in./cycle
R

io-5

OG_(

a_

Symbolo
-I

-IO
0

Oo

io-6

0
0

-I0
0

0
-7

I0

I0

15
Ak,

20

25

30

35

ksi-in

1/2

Figure 6.- Variation of fatigue-crack-growthrate with Ak for R = 0 and R = -1 in 2024-T3alloy.

23

Ak

,MN/m

I0

iO-I'

3s2

20

Ak
30

4o o

0
R_<O

10 -2

,MN/m

_o

A k,M N/m 3/z

3/2

30

29

40

,o

zo

_o

40

lo

O'

o
/

R=0.33
AA

o00O

,o5
Role,
nm/cycie
i

iO -_
Role,
in./cycle

10 -4
3
I0
iO -5

,d
10 -6

iO -I

,o"
R=O.5

R=0.7

10 -2

R--O.B

io_

i0 -_

Role
nm/cyc

,o"

Rate,
in./cycle
10 .4

,o3
iO -5

,o'
'-'o

i
,o ,5 20
_s 3'o _:so-_--%

Ak,

ksi-in

,'s 2'o 2'5 s'o s'so

_/2

Ak,ksi-in

(a)

Figure

24

l.-

Correlation

of experimental

fatigue-crack-growth

7075-T6

rates

w2

I
i
I
,o
is
20
J5 3'o _5
Ak,

ksi-in

V2

alloy.

at various

stress

ratios

with

Forman's

equation

(ref.

2).

Ak,M
IO

10-3

Ak,MN/m

N/rn 3/2
20

R=O

30

IO

400

3/z

20

R=0.33

,o'
Rate,
nm/cycle

Id"

'03

Rate,
in./cycle

,O_
Id _
IO

R = 0.7

R=O.5

,o_
Rate,
nm/cycle

Id"
Rate,
in./cycle

iO-5

io_

IO

I0

15

20

2'5

30

3'5

Ak,ksi-in
(b) 2024-T3alloy.
Figure 7.- Concluded.

25

,Ak ,MNlrn
I0

iO-I

3/2

20

Ak ,MN/m
400

30

I0

3/z

/k k, MN/m 3/2

20

3O

40

I0

20

4O

30

IO
o

R_O

IO-2

R=O2
o_

R=O.33

I0

Rote,
iO--_

nm/cycle

Rote,
in. / cycl_

I0

iO-_
Erdog

an's

equation

-_-

IO

IO-"
2

IO
IO

IO

IO-I1
R=0.7
i0"_ I

R=O.8

R = O. 5
5

IO
o

i0-_

Rate,
nm / cycle

Rote,
in/cycle

IO

10-4

io'
10-5
2

IO

/
Id60

I0

15

20

Ak,ksi-in

2'5

3;0

V2

3'5 0

IO

1_5-- 2_O
Ak,ksi-in

2'5

3_0

3'5

IO

I/2

15

20

A k,ksi-in

25

3'0

3'5

I/2

(a) 7075-T6 alloy.

Figure

26

8.-

Correlation

of experimental

fatigue-crack-growth

rates at various

stress

ratios

with

Erdogan's

equation

(ref.

3).

I0

R=O
0

10"4
Rate,
in./cycle

iO-S

10-6

Erdogan's

equation_

R=0.5

Id 4
Ra re,
in./cycle
I0""

_k, ksi.inW2

2?

tO

Ak,MN/m
_/2
20

30

40

Ak, MN/m 3/2

20

I0

Ak,MN/m 3/z
20

I0

400

30

3O

ooi
R_<O

oO_O

R=02

R=033

'05

AA

Role,
nm/cycle

A &

10-3

,o_

Rate
n/cyc e

paris'

equati_on

10-4

lO -_

,o_
,

10

[o'
lO-I I
R =0.8

R=07

R=05

,o_

I0"21

Rote,
nm/cycle

0
i0 -3
Rote,
in/cycle
10 -4

'

'03

I0 -_

iO -_

I0

15
20
Ak,ksi-in v2

_5 s'o s'so

I0

;'s _'o _5 _o s_o

Jo l's _o _5
Ak ,ksi- inm/_

Ak,ksi-inv2

(a) 7075-T6 alloy.

Figure 9.-

28

Correlation

of experimental

fatigue-crack-growth

rates at various

stress ratios

with

Paris'

equation (ret.

4).

I0

_2 q

Z_k,MN/m 3/2
20

400

30

Ak,MN/m 3/2
20

I0

30

I0

R=O

R =0.33
4

I0
Z_

Rate,
nm/cycle

I(/4
Rate,
ir_/cycle

I0

165
I0

10.6
I0

Id

i0-_

'05
R=0.7
4

I0
Rate,
nm/cycle

Id'

Rate,
in./cycle

I0

Ig 5

(b) 2024-T3 alloy.

Figure

NASA-Langley,

1969

--

]7

L-6662

9:

Concluded.

29