NO L BONDED 30DM/1010 STEEL METALMETAL AMINATES 1 ,

r ~l.A A A
I7
-111I FORG, EAB I TOUGH NESS..U) VOUGHT CORP ADVANCED
TECHNOLOGY CENTER DALLAS TX E 5 0 ER JUN 82

Lflflflllffff~ll
mommEmmohmEHE
IIg
111111.05 I1

REOUTO TES22AR
MICRCOP
LBUEUOSTNAD-16A
NA)N

ANi
 AD A 12 ' *- ±
CONTRACT NO. N00019-80-C-0575
ATC REPORT NO. R-92000/2CR-20

Roll-Bonded 300M/1010 Steel Metal-Metal Laminates:

Forgeaility, Toughness, Fatigue, and Stress Corrosion

L. E. SLOTER

VOUGHT CORPORATION ADVANCED TECHNOLOGY CENTER

P. 0. BOX 226144
DALLAS, TEXAS 75266

JUNE 1982

FINAL REPORT FOR PERIOD SEPTEMBER 1980 - DECEMBER 1981

APPROVED FOR PUBLIC RELEASE; DISTRIBUTION UNLIMITED

Prepared for:

DEPARTMENT OF THE NAVY

OTIC
ELECTE
NAVAL AIR SYSTEMS COMMAND MAY 0 2 MD
lo-WASHINGTON, D. C. 20361
m CL

> 000.

-J VOUGHT CORPORRTIOn
LA. ho
1clvancuc Tech loqL caotlr

83 05 02 01'Z
__
* UNCLASSIFIED
SECURITY CLASSIFICATION OF THIS PAGE (fhin Data E01104
9

REPORT DOCUMENTATION -PAGE as" Sa a

IREPORT NUM1111R 2.GOVT ACCESNN.7 W
EI11I' AALOG N R

14, TILE (ad Subitle)S. TYPE or REPORT & PtUSoo CalOVINE

Roll-Bonded 30014/1010 Steel tMetal-!4etal Laminates: Final Report for Period
Forgeability, Toughness, Fatigue, and Stress 18 Sept 1980 - 18 D~ec'1981
Corrosion 6. PmpoAINue Oe. sweaRT Numeaf
_______________________________________R-9200042,CR-20
7. AU THOR(*) 0. CONTRACT OR GRANT 1111116911(d

L. E. Sloter N00019-80-C-0575
9. PEarPO"UIt" ORGANIz ATION NAME AND ADDRESS PS 51 M Is
%f@IA: M T'AIK
Vought Curporation Advanced Technology Center 329 N u

P.D. Box 226144

Dallas, Texas 75266
11. CONTROLLING OFFICE NAME AND ADDRESS I.REPORT OATS
Department of the Navy June, 1982
Naval Air Systems Command is. Mumma" or PAGES
Washington D. C. 20361 97
IC. -MONITORING. AGENCY NAME 4 ADORESS(if diffeet from, CaeitrofiI~ Offics) IIS. SECURITY CLASS. (of le. repes

UNCLASSIFIED AN

1S. OISTRIGUTION STATEMENT (of this Report)

Approved for public release; distribution unlimited

17. DISTRIBUTION STATLoMENT (of the abstract .entered In &1eac 20. If different has.ReSect

IS. SUPPLEMENTARY NOTES

It. KEY WORDS (Continue on revere aide inecayan~ud ldeni& by. block .embor)
Alloy 300M4, Fatigue Strength, Forgeability, Forging, Fracture Toughness,
K ,Laminates, Metal Laminates, Plastic Buckling, Roll'Bonding, Steel,
SIMS'Corrosion Cracking

20. A&STRACT (Cotinhe an revorae side It necesary and IdetiE by bleak am""s
The fabrication of Alloy 30014/1010 steel metal-metal laminates in thick section
by hot roll bonding is described and discussed. The results of the forgea-
bility testing of the roll-bonded laminate plate and the tensile, fractute
toughness, fatigue, and stress-corrosion cracking properties of the roll-
bonded and roll-bonded and forged laminate material and 30014 monolithic control
material are reported, evaluated, and discussed. Two thick laminate billets
in which eighteen layers of 30014 low alloy steel separated by seventeen inter-
leaves of 1010li
DO,1j" 1473 EDITION OF I NOV 6SI1OS0SOLM
ETESIFr
S/N 0102- LP.F--6601 CVj-VL%"M*A OTwOgfibae
 UNCLASSIFIED
SECURITY CLASSIFICATION OF THIS PA,, (Wq D "WOOte

and sheet stock. Cylindrical specimens machined from these billets have been
forged in both open and closed dies and the laminate forgeability established
The forgeability of the laminate in the longitudinal direction, i.e., mater-
ial flow perpendicular to the lamellar plane, is limited to low strains,less
than -27%, by the hot strength of the interlamellar bonds and the plastic
buckling resistance of the laminate. The forgeability in the transverse
direction, i.e., material flow parallel to the lamellar plane,is limited only
by the acceptable reorientation of laminate fiber caused by die shape and
the frictional characteristics of the process, the transverse flow of thp
layers and interleaves being completely stable in the 300M/1010 system. 7
effects of forging on the mechanical properties of the laminate primarily re-
sult from the decreased layer thicknesses and improved bond integrity that
obtain with flow. Improved fracture toughness results from the decreased
layer thickness up to a maximum at which the improved bond integrity begins
to degrade the ability of the layers to fracture independently. An improve- L
ment in toughness of 300 percent over monolithic 300M is evinced by a forged
laminate. The tensile strength of the laminate is decreased with respect to
monolithic 300M by the volume fraction of low strength mild steel in the
laminate and by the plastic constraint imposed on the layers by the interleavis
during tensile elongation. Tensile strengths for the laminate materials
were above 200 ksi (1379 MPa) versus 263 ksi (1813 MPa) for the monolithic
300M. The laminate uniform elongation and tensile elongation are superior
to the monolithic material. In high cycle fatigue tests the superior
damage tolerance of he roll-bonded laminate results in improved fatigue
strength at least 10 cycles. Finally, the stress intensity factor for
stress-corrosion cracking, KIscc, in simulated sea water is significantly
increased in the roll-bonded la inate system versus monolithic 300M, being
32 ksi-inl/ 2 versus 15 ksi - in1/ 2 (35 versus 17 MPa - mi/ 2 ). The stress-
corrosion crack propagation rates at high stress intensities are approximatel
equal in the laminate and monolithic systems.

i NCLSSI F E
S/N 0102- Lfk 014- 6601
UNCLASSIFIED
89CURITY CLAWFICATIO. OF T1S PAGUM.. D.* E,0
 V

PREFACE

This report is a description of the procedures and results of the "Forged

Laminates for Structures Study" conducted at the Vought Corporation Advanced
Technology Center during the period 18 September 1980 through 18 November
1981. The study was conducted for the Naval Air Systems Command under Con-
tract N00019-80-C-0575. The contract monitors for the study were Mr. Michael
Valentine and Mr. Joseph Bruce, Naval Air Systems Command, Codes AIR-5304B4
and AIR-5304B5, respectively.
The study was conducted under the supervision of Dr. D. H. Petersen, Man-
ager - Structures and Materials Research; the Principal Investigator was
Dr. L. E. Sloter. Commendable technical support was provided by Messrs.
J. H. Thomas. F. H. Ebel, B. K. Austin, and C. P. Robinson. Additional techni-
cal support was provided by Dr. L. J. Cuddy and Messrs. R. C. Adams and
J. C. Raley of the United States Steel Research Laboratories, Monroeville,
Pennsylvania, in the roll bonding of the steel laminates, by Dr. S. C. Jain of
the Beaumont Well Works Company, Houston, Texas, in the open die forging of
the laminate, and by Messrs. W. E. Latta, P. G. Evers, J. Pillar, D. Leitgeb,
T. Bain, and T. Egnot of Pittsburgh Forgings Company, Corapolois, Pennsyl-
vania, who were all extremely helpful in the arrangement and perfomance of
the closed die forging of the laminate material on standard production equip-
ment. In addition to the technical suggestions of Messrs. Bruce and Valen-
tine, the interest shown and the suggestions contributed by Mr. R. Schmidt,
Code AIR-320, and Mr. J. F. Collins, Code AIR-5304B, have been very valuable
to the nature and content of the study.

Accession For
NTIS A&
DTIC TAB
Unannounced
IJustification' Q3
By on/ !

Distribution/
Availability Codes
jvail and/or
lo~t Special

) .

S--(~
 TABLE OF CONTENTS
PAGE NO.

PREFACE

1.0 INTRODUCTION 1

2.0 EXPERIMENTAL PROCEDURE 10

2.1 MATERIAL SELECTION 10
2.2 LAMINATE PREPARATION 10
2.3 FORGEABILITY TESTING 13
2.4 COMPLEX FLOW EVALUATION 15
2.5 PHYSICAL EVALUATION 17
2.6 HEAT TREATMENT 20
2.7 MECHANICAL TESTING 20
2.7.1 Tensile Properties 20
2.7.2 Fracture Properties 20
2.7.3 Fatigue Strength 23

2.8 STRESS-CORROSION CRACKING EVALUATION 23

3.0 RESULTS AND DISCUSSION 24
3.1 LAMINATE PREPARATION 24
3.1.1 Laminate Micrography 24
3.2 FORGEABILITY TESTING 26
3.2.1 Tensile Properties 41
3.2.2 Micrography 45
3.3.3 Fracture Properties 49
3.3 COMPLEX FLOW EVALUATION (CLOSED DIE FORGING) 56
3.3.1 Micrography 58
3.3.2 Tensile Properties 62
3.3.3 Fracture Properties 65

3.4 GENERAL DISCUSSION OF LAMINATE FRACTURE TOUGHNESS 68

3.5 GENERAL DISCUSSION OF MATERIAL FLOW AND FRACTURE TOUGHNESS 72
3.6 FATIGUE STRENGTH OF ROLL-BONDED LAMINATES 75
3.7 STRESS-CORROSION CRACKING OF ROLL-BONDED LAMINATES 80

4.0 CONCLUSIONS 85

5.0 RECOMMENDATIONS 87
REFERENCES 88
DISTRIBUTION LIST

iv
.1 I
 1.0 INTRODUCTION

The practice of utilizing layers of dissimilar materials in order to

achieve desired properties in structures and other useful items originated
during prehistory. By the time the early Middle Eastern and Uriental civili-
zations were firmly established the exploitation of layered materials for both
aesthetic purposes, such as wood veneer and gilding, and structural purpQses,
such as the banding of wood with metal, already had become quite sophisti-
cated. 1 During the past several decades the application of highly advanced
laminar composites has been effected, particularly in the aerospace industry.
Such laminated composites offer the potential for improved reliability, dura-
bility, and damage tolerance combined with lower maintenance requirements and
life cycle costs when compared with conventionally wrought and hmachined
items. Other appealing properties of laminates are the high specific moduli
and strength characteristics and higher fracture and fatigue resistance poten-
tially achievable through proper laminae and lamination process selection. In
particular, the enhancement of fracture toughness and fatigue resistance
through lamination has been the subject, of intense investigation. 2 "2 8 Lam-
inates Investigated have included single constituents bonded together 2-25
j and duplex or multiplex laminates formed by a variety of techniques. 2 6 -36
Although these studies have shown that lamination is beneficial for fatigue
* and fracture resistance, only the investigations of hlson19 and Goolsby 4
have provided a thorough evaluation of lamina and laminate thickness effects
on fracture toughness, using current fracture toughness testing procedures.
In particular, the early work 4 conducted at the Vought Advanced Technology
Center has shown that laminate toughness can be maximized by proper selection
r -: of lamina material and thickness.
Metal laminates are a type of laminar composite and typically consist of
multilayers of bonded homogeneous laminae as shown in Figure 1. Although the
layers are generally of a single homogeneous material, each layer aiay be iso-
tropic, orthotropic, or anisotropic depending on material properties, such as
the basic crystalline structure and texture. The type of bonding provides a
convenient classification of laminar composites, the two basic bonding techni-
ques employed involving either a direct metal-to-metal (metallurgical) bond or
 BONDING PRIMARY
LAYERS METAL
LAYERS

FIGURE 1. SCHEMATIC METAL LAMINATE.

2
 an adhesive bond, such as that produced by an epoxy adhesive layer. Bonding

with adhesives is a relatively easy low temperature process that provides the
necessary conditions for high fracture toughness. Metal-to-metal bonding
typically is achieved by diffusion bonding, brazing, soldering, roll bonding,
or explosion bonding and may or may not provide the necessary delamination
conditions required for high fracture toughness.
Previous investigations 4 ,3 3 -3 6 performed at the Vought Corporation
Advanced Technology Center (ATC) have established that significant benefits
are derivable through the use of metal-metal laminates. These laminate mater-
ials offer high toughness, crack arrest capacity, and subcritical crack growth
resistance. The ATC Metal Laminates for Structures Program has demonstrated
these beneficial capacities in plate materials formed by explosion bonding,
diffusion bonding, and roll bonding. In particular, roll-bonded metal-metal
laminate plates composed of layers of high strength metal alloys and inter-
leaved with dissimilar lower strength alloys (Figure 2) have been shown to
have thick section tensile and fracture toughness properties that are superior
to a corresponding section of monolithic or unlaminated alloys. These super-
ior properties derive from both the crack dividing and the crack arresting
properties of the generally softer interleaf alloy. Since the crack arrest
and crack divider orientations are of importance in orienting laminate frac-
ture properties and laminar composite application, they are illustrated in
Figure 3. Mechanistically in the crack divider orientation, the controlled
delamination of the layers during the fast fracture of thick laminates results
in a plane or nearly plane stress rather than a plane strain fracture, and,
therefore, the energy required for fracture or the fracture toughness is
greatly increased. This rationale is illustrated schematically in Figure 4 in
which the general inverse relationship of fracture toughness versus thickness
is contrasted with laminate behavior. As shown in the figure, the toughness
of most materials decreases from a plane stress maximum at thin section sizes
through a mixed stress region to a pure plane strain limiting value that may
be defined as the plane strain fracture toughness. 37 In the crack arrest
orientation, the interleaf effectively blunts a running crack by plastic flow
in the case of a soft interleaf or interfacial separation, thereby slowing or
stopping the crack and concomitantly improving the fracture toughness. More-
over, laminates may be fabricated to provide other enhanced properties, such

_ __k~
 LUJ

-LJ

LL.LU

J -LJ

LA-

LLI)

ci\J

LLiJ

C/

~~ I-
Iow

FIGURE 3. ()CRACK ARREST AND (b)CRACK DIVIDER ORIENTATIONS.

1 5

=17 -.--
 -LJ

LI

00
uii
-

-)

0i
wC
CL C

-LJ

0
L6C I 0

I E
cw
Wi U
V)

I CD

Li

/ LIl
LL.

6
 as mechanical and chemical stability, and improved wear resistance and cor-
rosion resistance through the proper choice of layer and interleaf alloys, the
geometry of the laminate lay-up, and the processing and heat treating vari-
ables.
In previous years of the Metal Laminates for Structures Program the
metal-metal laminate concept has been validated and various fabrication tech-
niques explored. During the first year of the program, seven laminate config-
urations were fabricated using three processing techniques; diffusion bonding
roll bonding, and explosion bonding. The materials systems investigated were
7475 A1/1100 Al (the alloy designated to the left of the slash mark is the
primary or layer alloy; the alloy to the right, the interleaf), 7075 A1/7072
Al, and Ti-6Al-4V/6061 Al. The mechanical properties - strength, fracture
toughness, and fatigue strength - of each laminate system were evaluated and
compared with similarly heat treated monolithic alloys.4 During the second
year of the program, diffusion bonded 7475 A1/1lO0 Al, 7475 Al/ll00 Al,
7075 A1/7072 Al, Ti-6A1-4V/CP-Ti, ultrahigh carbon steel/interstitial free
iron; adhesively bonded 7475 Al and 7075 Al; and roll bonded 7475 A1/1100 Al
were evaluated similarly.3 3 In the third year the further development of
roll bonding procedures and heat treating parameters was effected for several
steel and titanium alloy systems. Specifically studied were 300t1/AISI 1020,
30OM/SAE 1075, and 300M/AISI E52100, Ti-6A1-4V/CP Ti, Ti-lOV-2Fe-3A1/Ti-l5V-
3Cr-3AI-3Sn, Ti-8A1-eV-6Cr-4Zr-4Mo/CP Ti, and Ti-8Al-8V-6Cr-4Zr-411o/Ti-15V-
3Cr-3AI-3Sn. In addition to the tensile and fracture toughness of these alloy
laminate systems, the axial fatigue properties of representative alloys were
evaluated. Finally, the use of several interleaf alloys in the 300M steel
system allowed the conjoint nature of interleaf properties, interlaminar bond
properties, and layer properties to be evaluated, and the inverse relationship
between layer thickness and fracture toughness was substantiated. 3 5 Some
results for several laminate systems investigated in the Program are listed in
Table 1.
Although many items may be fabricated directly from plate, numerous other
structural components benefit from the material utilization efficiency and
forming economy occasioned by forging. In addition, the mechanical properties
of many components may be improved by the flow or fibering that obtains during
forging. Nevertheless, the forgeability of laminates and the effects of forg-
ing on roll-bonded metal-metal laminates had not been determined. Further-
more, laminates present several unique problems in forging since the improved
l7
jr ;
JJLJ cvli ,

t~,~.- -

L"~ ~ 30C)~ 0e
CC -

&A 00 -NJ.

V~) NJ - - - -

CD on 00 0
Cl

LU

.. J C.) - -

V)* .r19 -,

LU -L. --

-3 LA- 0.J L0L . .-

LUL

I0-%- - C in

cr- .
LLa N~.J Cl~ - .

040

z1 C
0--V L. )
0~ IV -t
V)-V

CLene
1 ------
--------

fracture toughness of laminates is dependent on delamination and the fiber of

laminates is intrinsic and cannot be easily reoriented. The present study,
therefore, was undertaken to determine the forgeability of a roll-bonded
metal-metal laminate system in several orientations and the effects of forging
on the structure and properties, especially the fracture toughness, of metal
laminates. In addition, the stress corrosion cracking properties of roll-
bonded material were to be determined, since it had been suggested earlier
that these laminate material systems would offer significant benefits in
stress corrosion cracking resistance in the crack divider orientation. Spe-
cifically, the following tasks were perfomed for the study:

o Metal-metal laminate plate stock of 30014 alloy steel interleaved with

1010 mild steel was fabricated by hot roll bonding.

o Specimens taken from the plate stock and monolithic controls were
forged in an open die. The laminate was forged in both the longitudi-
nal and transverse orientations.

o The physical and mechanical properties of the forged laminates and con-
trols were determined when possible.

o Several laminate specimens were forged in a more complex closed die and
the mechanical properties of these forgings determined.

o The high cycle fatigue strength of roll-bonded laminate plate was com-
pared with monolithic control plate.

o The stress corrosion cracking rate of roll-bonded laminate plate and

the critical stress intensity factor for stress corrosion cracking were
determined and compared with a monolithic control.

h9

I
 2.0 EXPERIMENTAL PROCEDURE

2.1 MATERIAL SELECTION

During the execution of a prior stuy 35 the significant toughness Im-

provements obtainable in an ultrahigh strength steel/mild steel laminate sys-
tem fabricated by roll bonding were demonstrated. The most successful steel
system studied consisted of layers of 3004 alloy steel (MIL-S-8844C, 3 8 Class
3) interleaved with AISI-SAE 1020 mild steel. Since 30U14 is a coi~on aero-
space forging alloy that is used at high strength levels in critical applica-
tions, such as aircraft landing gear and arresting gear, a laminate system
based on 30014 was determined to be a desirable choice for the study of laui-
nate forgeability. Furthermore, a significant data base on roll-bonded 3UOH
systems had already been accumulated and could be used for comparison. The
specific material combination chosen consisted of eighteen layers of 3UUJ1 in-
terleaved with seventeen layers of AISI-SAE 1010 mild steel.

2.2 LAMINATE PREPARATION

A 3/4 inch (19 m) plate of steel alloy 3004 (MIL-S-8844C, Class 3) weigh-
ing approximately 390 pounds (177 kg) was obtained from Friend Meetals,
Anaheim, California. The mill chemical analysis of this plate is listed in
Table 2 in which it is also compared with the specified chemical analysis.
Since a one-fourth inch (1/4 in., 6.35mm) starting thickness was required for
the laminate lay-up prior to roll bonding, plates of 6 x 8 inches (152 x
203amm) dimensions were flame cut from the parent plate and rolled to the one-
fourth inch (6.35 m) starting thickness required at the Jones and Laughlin
Steel Corporation Graham Research Laboratories, Pittsburgh, Pennsylvania. A
start temperature of 2150OF (1177 0 C) was used for the rolling, and the re-
duction of 67 percent was accomplished without reheating. Following rolling,
the 30014 plates were cut into approximately 17 inch (432 m) lengths, grit
blasted, degreased, and laid up with interleaves of coiiercially obtained
AISI-SAE 1010 mild steel (Table 3) sheets approximately 0.030 inch (U.76 ti)
, thick to form an eighteen layer, seventeen interleaf metal-uaetal laminate
lay-up, approximately 5.35 inches (136 m) in height. Finally, the lainate

10
 TABLE 2. CHEMICAL COMPOSITION OF ALLOY 3001

CONTENT, WEIGHT PERCENT

ELFE AT SPECIFIED*4 0 MILL ANALYSIS

C 0.40-0.45 0.420

Mn 0.65-0.90 U.74

Si 1.45-1.80 1.73

NI 1.65-2.00 1.83

Cr 0.65-0.90(0.70-0.95) 0.86

No 0.30-0.45(0.35-0.45) 0.41

V 0.05 MIN 0.09

P 0.025 MAX (0.010 MAX) U.006

S 0.025 MAX (0.010 MAX) 0.004

Cu --- 0.16

Fe Balance ---

*HIL-S-8844C requireilents which differ fria commercial 30014 are listed in

parentheses.

11

___
 TABLE 3. CHEMICAL COIPOSIT ION OF
AISI-SAE 1010 AND 4145 STEELS 4 1

ELEMENT CONTENT, WEIGHT PERCENT

1010 4145

C 0.08-0.13 0.43-0.48
IPin 0.30-0.60 O.75-1.00
P 0.040 MAX 0.035 MAX

S 0.050 MAX 0.040 MAX

Si 0.60 14AX 0.15-0.30

Cr --- 0.80-1.10
14o --- 0.15-0.25

Cu 0.060 MAX ---

Fe Balance Bal ance

12
 lay-up was boxed in a 1010 mild steel box and the box welded closed. The pur-
pose of the box was the stabilization of the laminate during the initial roll-
ing prior to complete bonding and the protection of the surfaces during heat-
ing and processing. The boxed laminate lay-up was roll bonded by hot rolling
using a start temperature of 2150°F (1177 0 C) and an overall reduction in
area of 44 percent. The rolling was accomplished without reheating using roll
reductions of approximately 10 percent per pass and six passes on a 1O1)0 kip
(4.5 MN) reversing mill at the U.S. Steel Corporation Research Laboratories,
Monroeville, Pennsylvania. The final thickness of the roll-bonded laminate
was three inches (76 m). Bonding was complete in the center of the laminate
billet, although there was a central delamination at each end. The delamina-
tion extended approximately three inches (76 mm) into the billet at one end
and one Inch (25 mm) at the other. The material above and below the plane of
delamination was bonded, however. The as-roll-bonded laminate billet follow-
ing the removel of the delaminated end is shown in Figure 5. The cut end has
been macroetched with dilute nitric acid to reveal the layered structure.
)
2.3 FORGEABILITY TESTING

The forgeability test chosen for the laminate and control material was the
upset test. 3 9This procedure consists of upsetting identical cylinders of
material to varying thicknesses or cylinders with differing height to diameter
ratios to the same final thickness. Since it was desired to use the full
thickness of the laminate billet and to decrease the relative effects of fric-
tion through the constancy of initial contact area, the latter method of upset
testing, viz., the varying height to diameter ratio, ho/do, method was
chosen for the longitudinal and control forgings. The choice of this method
also facilitated the mechanics of testing in that a single stop block could be
used for all the tests, this method being the only viable one for controlling
thickness in the large hydraulic open die press used. The limited laminate
billet thickness, however, necessitated the upsetting of flat plates of mater-
ial to varying heights for transverse upset testing, but the small relative
change in surface area during forging did not pose any problems vis a vis
friction. For the longitudinal forgeability testing the laminate billet was
cut into three inch (76 m), four inch (102 m), and five inch (127 m) blanks
and these were turned to form cylinders 2 7/8 inches (73.03 u) in diameter

*13
vz

FIGURE 5. AS ROLL-BONDED 300M/IO10 LAMINATE BILLET. THE CUT

END HAS BEEN MACROETCHED WITH NITRIC ACID. THE
BILLET DIMENSIONS AS SHOWN ARE 2 7/8 x 6 3/8 x 13 5/8
INCHES. (73 x 162 x 346 mm)

I _-14
 and of the following nominal lengths: 2.5 In. (64 m), 3 in. (76 um), 4 in.
(102 mm), 4 in. (102 m), 5 In. (127) m, and 5 in. (127 mm). In addition,
comparison cylinders were turned from a four inch (102 mm) square rolled bil-
let of AISI-SAE 4145 low alloy steel such that the axis of the cylinder cor-
responded to the rolling direction of the parent billet. The specified com-
position ranges for 4145 are listed in Table 3. The 4145 cylinders were also
2 7/8 Inches (73 mm) in diameter and, respectively, 3 in. (76 mm), 4 in.
(102 mm), and 5 in. (127 m.) in length. In all cases particular attention was
paid toward the parallelism of the machined cylinder ends. One of the as-
turned laminate cylinders is shown in Figure 6. In addition to the cylinders,
four transverse forging specimens were prepared of the following dimensions in
inches (umm): 6.250 x 3.750 x 1.375 (158.75 x 95.25 x 34.93), 6.688 x 4.125 x
1.875 (169.88 x 104.78 x 47.63), 4.250 x 2.625 x 1.332 (107.95 x 66.68 x
33.83), and 4.250 x 2.625 x 1.616 (107.95 x 66.68 x 41.05). In this and fol-
lowing discussions longitudinal forging refers to processing In which the di-
rection of ram travel lies in the laminar plane of a lainate specimen or is
concident with the rolling direction of the monolithic billet; the forging
plane is normal to the laminar plane or monolithic fiber. Transverse refers
to the inverse orientation of the laminate, i.e., the ram direction is perpen-
dicular and the forging plane is parallel to the laminar plane.
The upset forging was performed on a 3300 ton (29.4 N) I'draulic press at
the Beaumont Well Works ComparY, Houston, Texas, using flat, parallel open
dies and a start temperature of 2200OF (12049C). Lubrication was provided
by woven glass cloth placed between the hot specimen and the dies. Forging
speed was in the range of 0.25 - 1.0 ft/s (0.08-0.30 m/s). Final height of
the longitudinal cylinders was controlled to between 2 and 2.5 inches (51 and
64 mm) nominally and to between 0.5 and 1 inch (13 and 25 m,) for the trans-
verse specimens.

2.4 COMPLEX FLOW EVALUATION

Although the open die testing was designed to provide the requisite infor-
* mation on forgeability in simple flow geometries, it did not simulate the more
complex material flow often required in the forming of a complex structural
item. Therefore, a closed die forging of more complex shape was chosen to
supplement the open die forgeabilitv testing. In order to minimize the costs

15

K ... .1.17117.7 -
 (b)

j __ _
FIGURE 6. LAMINATE LONGITUDINAL FORGEABILITY CYLINDER.

__ ____16
 r
associated with the closed die forging, it was necessary to use an existing
production die and, of course, standard production equipment. The die chosen
was that for a gear blank. The part produced from this die resembled a wheel
and hub. The forging print for this part with flash removed is reproduced in
Figure 7. As may be noted in the figure the item consisted of a central hub
surrounded by a circular web and a circumferential flange.
The roll-bonded stock material for the closed die forging was fabricated
identically with that used in the open die forgeability tests, that is, a
second billet was laid up and roll bonded as described in Section 2.2. This
billet was sectioned and the blanks turned to produce six cylindrical preforms
approximately 3 1/4 inches (82.6 mm) in diaseter and 2 7/8 inches (73 mu) in
height. The presence of a large lamination In the billet, however, prevented
the fabrication of more than one preform of this height, the other five being
approximately 1 7/8 inches (48 mm) in height. Although these later five pre-
forms did not throw flash, they did very nearly fill the die and did form the
outer flange. The process schedule for the closed die forging is sumaarized
schematically in Figure 8. The forging itself was perfoned on a board haumer
at the Pittsburgh Forgings Company Coraopolis, Pennsylvania, plant. The pre-
forms were heated in a gas fired slot furnace to approximately 2400°F
(1316 0 F) and then forged, the full height prefom requiring 27 blows and the
remaining five preforms eight or nine blows each. Following forging each of
the material specimens were annealed and either sectioned and examined or
machined for mechanical testing.

2.5 PHYSICAL EVALUATION

Following forging each forgeability specimen was visually examined then

subcritically annealed at 1275°F (6910 C) for two hours, the AcI for 3UaM
being approximately 1400°F (760°C), in order to facilitate the machining
of mechanical test specimens from the forgings. Of course, prior to machining
the forgings were photographically documented and examined or plvsical integ-
rity. Samples were also taken from the forgings for metallographic examina-
tion, particular attention being paid to the layer-interleaf interface and its
I nteg ri ty.

17

* i _ _ _ _
 -AJ

LJ

C3

LUJ

rj- X

Lj-

LU

LL.

18~

.1 'n4 __"'A
 D,

CDV
U-

CD J(

060

-In

C)W

19J
 2.6 HEAT TREATMENT

All the specimens were heat treated so as to produce strength levels that
were near the practical maximum for 30014 steel, since the value of metal lami-
nates has been shown to be primarily in the retention of fracture toughness at
ultrahigh strength levels. The specific heat treatment schedule used was the
following: normalize at 1700°F (927C) for 1/2 hour (1.8 ks), air cool to
approximately 400 - 500°F (204-2600C), austenitize at 1600OF (8710 C)
for 1/2 hour (1.8 ks), oil quench, temper at 575°F (3U20 C) for two hours
(7.2 ks), air cool, and repeat temper. This heat treatment produces an ulti-
mate tensile strength for 3004 greater than 250 ksi (1724 14Pa).

2.7 MECHANICAL TESTING

Following heat treatment the mechanical properties of those forgings from

which sufficient mechanical test specimens could be machined were evaluated
through tensile and fracture toughness testing. In addition, the important
physical properties of the specimens, such as the thicknesses of the individ-
V4 ual layers and interleaves and the average layer thickness were determined.

2.7.1 Tensile Properties

f
Tensile specimens were cut from the as-rolled and the forged material
where possible in accordance with American Society for Testing and Materials
(ASTM) Standard E 8 for plate. 4 2 Tensile testing was accomplished in accor-
dance with ASTM E 8, and the following tensile properties were determined:
the 0.2 percent offset yield strength, the ultimate tensile strength, the ten-
sile elongation or elongation to fracture, the uniform elongation, the true
strain at the onset of necking (maximum load) or the true uniform elongation,
and the true stress at necking (maximum load).

2.7.2 Fracture Properties

The fracture toughness of laminates and monolithic materials was deter-

mined from standard, ASTM E 399,43 compact tension specimens. These speci-
mens were machined from the annealed material, heat treated, and then pre-
cracked in accordance with ASTh E 399. Specimens were machined from the plate
20

i I I. -I- - AIN
 and forgings in as many orientations as were compatible with the processed
material. All specimens were of crack divider type, since this is the tough-
ness controlling orientation expected in most laminate structural items. Fig-
ure 9 illustrates schematically the compact tension specimens orientation with
respect to the flow of the parent material, either plate or forging. Fracture
toughness parameters were calculated from measurements of the loads and the
corresponding crack opening displacements (CUD) experienced by the compact
specimens during testing. Three stress intensity factor toughness parameters
were calculated as follows:
o KQ - The conditional fracture toughness calculated using the 95
percent secant load (ASlM E 399) and the calculated (CUO)crack
length corresponding to that load.
o KA - the apparent fracture toughness calculated using the maximum
load and the same crack length as KQ.
o Kc - the critical fracture toughness calculated using the maximum
load and the effective calculated crack length (from COD) correspond-
ing to maximum load.
In all cases fracture toughness, KX, is defined as follows:
Pf
KX = B-1/2 f(a/W)

Pf = load at failure or crack extension,

B = specimen thickness,
W = specimen width, and
a = crack length at crack extension or failure.

In addition to the above toughness parameters, the specimen strength ratio, 4 3

2PMAx(2W+a)
RSC V=
B(W -0a y
.- y

where
PMAX = maximum load sustained
W - specimen width,
a = crack length,
B - specimen thickness, and
ay a the 0.2 percent offset yield strength

was calculated.
21
 WIDTH ROLLING DiIECTION
LONG TRANSVERSE DIRECTION OF MAXIMUM
DIRECTION OF
MINIMUM FLOW S
THICKNESS
SHORT TRANSVERSE

FIGURE 9. SCHEMATIC ILLUSTRATION OF CRACK PLANE ORIENTATION WITH RESPECT TO

THE ORIENTATION OF THE PARENT MATERIAL. AFTER ASTM E 399, REFERENCE 43.

t 22
 2.7.3 Fatigue Strength

The fatigue strength of roll-bonded laminates was compared with monolithic

300M through the generation of stress versus cycles-to-failure data. All
fatigue tests used cylindrical specimens machined such that the longitudinal
axis of the specimen corresponded to the rolling direction of the lainate
billet or monolithic plate. The fatigue tests were conducted in axial tension
with a load ratio, R = minimum load/maximum load, of 0.1 at a frequency of 3U
Hz in air.

2.8 STRESS-CORROSION CRACKING EVALUATION

The stress corrosion cracking properties; crack propagation rate, da/dt,

and the critical stress intensity factor for stress corrosion cracking,
Kiscc; were determined for roll-bonded 300M/1010 material in the crack divider
orientation and compared with monolithic 30014. These tests were conducted us-
ing bolt-loaded (self-loaded) compact tension specimens. Following heat
treatment these specimens were cleaned, precracked, degreased, loaded, and
placed in the 3.5 percent sodium chloride (NaCl) aqueous solution. Periodic
optical measurements were made to determine the crack growth as a function of
time. The Klscc of each specimen was considered to be that stress intensity
j factor at which no growth had been discernible for at least one hundred hours
(360 ks).

23

K __ _ _ .....
 3.0 RESULTS AND DISCUSSION

3.1 LAMINATE PREPARATION

As noted previously (Section 2.2) the ends of the roll-bonded billet

failed to bond at the center lamina. This problem has been encountered before
and appears to be primarily a result of tensile stresses generated because of
insufficient penetration of the plastic zone during rolling. This problem is
analogous to the alligatoring sometimes encountered in the rolling of
plate. 44 Of course, prior to the obtainment of inter-layer bonding, the
short transverse strength of the laminate is essentially zero and it cannot
resist the splitting. An additional problem also obtains in the roll bonding
of laminate lay-ups that can lead to or exacerbate the alligatoring phenome-
non. This could be called the flat die or seesaw effect and results from the
tendency of long plates compressed at one end to spread apart at the other,
that is to act as independent beams. This response is manifest in rolling
laminates primarily when the laminate thickness is small in comparison to the
roll diameter. In order to prevent its occurrence the lay-up must be stabi-
lized against tensile failure effectively in the short transverse direction.
In roll bonding of laminates ft is desirable, therefore, in addition to plate
stabilization and atmospheric protection, to provide short transverse tensile
hot strength especially at the front and rear end of the lay up. This may be
done by welding the laminate and providing sufficient weld metal at the ends
to carry the load, e.g., provide a single-U or -V groove weld of sufficient
depth between layers; strapping the ends of the lay-up with a doubler plate
either welded or clamped on; or boxing the laminate and strengthening the
front and rear ends of the box.

3.1.1 Laminate Micrography

.- The microstructure of the base 3004 steel is shown in Figures lW and 11 in

the quenched and tempered condition. The steel as shown is completely com-
prised of a very fine tempered martensite as expected for a through hardening
low alloy steel of this type. In addition, it may be noted that the micro-
structure of steel is very clean, that is there are few non-metallic inclu-
sions visible. This is also typical of a vacuum arc remelted steel of air-
craft quality.
24

---- A- _
 100 gm

FIGURE 10. 300M STEEL, QUENCHED AND TEMPERED, LONGITUDINAL

SECTION. MAGNIFICATION: 200X. 2% NITAL ETCHANT.

: 50 iim

FIGURE 11. DETAIL OF FIGURE 10.

MAGNIFICATION: 500X. 2% NITAL ETCHANT.

25
 Following roll bonding the laminate plate was subcritically annealed at
1275°F (6910 C) for two hours (7.2 ks) in order to produce a fully spher-
oldized structure that could be easily machined. The mircostructure of the
as-rolled and annealed laminate is shown in Figures 12, 13, and 14. Since the
300M transforms even on air cooling to martensite and bainite, 4 5 it may be
fully spheroidized by a subcritical anneal, i.e., held at a high temperature
below Ac1 . The fine structure of the 300H evidenced in the micrographs is a
very fine spheroidite plus some coarse tempered bainite. The ferritic-
pearlitic structure of the 1010 obtained upon cooling Is essentially unchanged
by the anneal. An interesting microstructural effect may be noted near the
bond line, however, in that there appears to be a spheroldized layer just
within the 1010 interleaf and a light etching layer adjacent to the bond line,
predominately in the 300M. It is felt that this "spheroldite free zone" re-
sults from the diffusion of alloying elements, especially carbon, into the
1010 during rolling thus allowing the layer within the 1010 to transfom on
cooling such that it will later spheroidize. The light etching layer, then,
would appear to be a zone depleted by the ripening of the spheroidal cementite
within the parent 1010 during spheroidizatlon. The kinetics and directional-
ity of both the transdiffusion and the depletion may be aided by the high sil-
icon content of the 30014. This is analogous to the under surface decarburiza-
tion that is sometimes noted in carburized 300M.46 Subsequent micrographs
of quenched and tempered material illustrate the hardenability of the inter-
diffusion layer in the parent 1010 interleaf. In the quenched and tempered
case, however, the depleted layer disappears, since the material transport
during tempering is not as long range as that during spheroidization.
Figures 13 and 14 also illustrate the as roll-bonded condition of the
layer-interleaf bond line. It may be noted, particularly in Figure 14, that
the metallurgical bond is quite extensive with some apparent recrystallization
across the bond line. The bond line is also quite clean, although there Is
some porosity distributed along the interfacial bond line.

3.2 FORGEABILITY TESTING

The numerical results of the hot forgeability testing are contained in

Table 4. It may be noted that engineering strains between approximately -23
percent and -55 percent were investigated for the longitudinal cylinders, and,

26

- m . I i i
 0.5 pm

FIGURE 12. 300M/1010 AS ROLL-BONDED LAMINATE, ANNEALED,

LONGITUDINAL SECTION. THE CENTER LAMINA IS
THE 9'1010
+ INTERLEAF."-' MAGNIFICATION: 44X.

K
+' :
2% NITAL ETCHANT.

_ _. . . '.I.

FIGURE 13. 300M/1010 AS ROLL-BONDED LAMINATE, ANNEALED,

TRANSVERSE SECTION. THE 1010 INTERLEAF IS AT
THE LEFT. MAGNIFICANTION: 20OX. 2% NITAL
ETCHANT.

27
 FIGURE 14. 300M/1O1O AS ROLL-BONDED LAMINATE, ANNEALED
LONGITUDINAL SECTION. THE 1010 INTERLEAF IS
I AT THE LEFT. MAGNIFICATION. 500X. 2% NITAL
ETC HANT.

1 28
 COl rn 00 0n C
vi-*
tCj 10 en
CNJ 0%
in~ (' 7C
-~l
Ch
P.-
t-" .-. LA Q ~CV)
()
LA
A
Cl

cc J ?%. 0%C CU C; A 0% 0 . -;

0 0 0 r 0 W 0 0 0 o 0 0 0

U- w% m O . O ( ~

V,

< en M 0% 0%
rl Cl r-- Ot. C
07% M Mr C
LA-(~ L A 0e LA %

ccLASE CD (D CD CD C) 01~ 5% l Co Co I In w
I-J Ln (V-
Lr n LA- LA- LA' CL-) (V-.
Ln . In
a% co V

2m 9% O - 09 r 9I- r,-.9 C9 9r,.

" C CJ IJ CV- CV) CV) CV) CV) .4 0% r- I-

I- 0u 0o 0c 0n 0 0 0 0 0 - L O C
(4 m V 5%.. LC
.,I) P- V V) t- cV) -1V -.* V d, A L 0 00 C %
D .

L I

LLJ0 cl (4 ** -O LA () 0% 0% cn LA0 LC--

-j - L 0 % aU V%. %C rV) V) D .Lne CO 0% '.0)P as 0 c 0a

<C -Ca LA
-V 00 .V CYN
LAr A 0 r -r t LA)
M LA
o -0 CO

4 .' -L - LC - -n - LO - O-.C -% nn

LL _jC COO 'fl. Or- LC') Ct.. CO LA

j L0
%fl %.n 0 c- C)0

.~~~'.j~~~~- * 0n C m.V M%C UV

e CLA
*
.Y)

L
m4LC *c )
~.C-.
%.aO
O
qO-cV
*n *LA
C
* V
'.Q
*LA
%Mt
LA
~V)
C)O
*-
C00 -
LA L%
cm.
*%
%0.
Ok
00A L

a,
7C *-l ~ - - U , CD -
0% - C C, 0- - ::C)-. D - -n cn 0- 0-' 0-D

_j - wt - - - n - Lc) S -c -

C A C COl0it O % CO C

jr-.0

OC
Z LA L '0 CO- -

06U 0 0 0 M 0 a a a 0 o

29
 between approximately -38 percent and -64 percent, for the transverse forg-
ings. In all cases some delamination or edge cracking occurred in the longi-
tudinal forging of the laminate cylinders, whereas no edge cracking or center
cracking occurred in the 4145 monolithic control cylinders. The transverse
laminates also were forged successfully. The upset cylinders exhibited four
post-forged fors as shown in Figures 15 through 19. Figure 15 illustrates
the stable barrelling of all the 4145 cylinders. No edge cracking or center
cracking was discernable. In Figure 16 is illustrated the least severe upset
test of a laminate cylinder in which barrelling has been accompanied by some
edge cracking. Nevertheless, many of the laminae remained bonded as shown in
Figure 17. Upon more severe deformation the laminates generally buckled in
one of two modes as shown in Figures 18 and 19. The first mode shown in Fig-
ure 18 was complete buckling of all laminae leaving a central hole through the
forged laminate. A second mode shown in Figure 9 was observed in several
specimens in which several of the center layers were stabilized by the buckled
outer layers and were upset almost stably, although sone local buckling in
each of the specimens within these center layers prevented the retention of a
good bond after forging. No cracking or bickling were encountered in the
transversely forged laminates, although some uneven flow due to friction ob-
tained as shown in Figures 20-22. The phenomena of barrelling and buckling
and the effect of friction will be discussed subsequently in greater detail;
however, several operational conclusions concerning the forgeability of these
metal-metal laminates are justified on the basis of the hot forJeability
testing. These conclusions are (1) the forgeability in the longitudinal di-
rection is limited to approximately -27 percent strain, (2) for specimens
forged to strains of approximately -53 percent and more the mode of failure
was hot buckling rather than edge cracking, and (3) the laminates may be
forged successfully to strains of at least -63 percent in the transverse
direction.
The hot upset testing and results discussed above represent the most
-. *severe forging conditions expected for a metal laminate in the longitudinal
orientation. The severity of upsetting is related primarily to the magnitude
of the frictional traction that acts on the specimen in contact with the die.
In all the upset tests a chemical reaction between the vitreous lubricant
cloth used and the forgeability specimen led to essentially a condition of
sticking friction. This means that there was no relative movement between the

30

LI_
 1O1imm

FIGURE 15. UPSET MONOLITHIC SPECIMEN 9A

EXHIBITING STABLE BARRELLING AND A
LACK OF EDGE CRACKING.

FIGURE 16. UPSET SPECIMEN 4A EXHIBITING EDGE

CRACKING.

31

___________________________________________ - ?i.
 10 mm

FIGURE 17. UPSET SPECIMEN 4A (FIGURE 16) DETAIL.

MAGNIFICATION: 2X.

.
i ____
l 32
 10mm

FIGURE 18. UPSET SPECIMEN 6A EXHIBITING HOT

BUCKLING.

FIGURE 19. UPSET SPECIMEN 6B EXHIBITING BOTH

HOT BUCKLING AND STABLE UPSETTING.

33
FIGURE 20. TRANSVERSE FORGED SPECIMEN 13, CROSS SECTION.
MACROETCHED IN DILUTE NITRIC ACID.

FIGURE 21. DETAIL OF THE CENTER CROSS SECTION OF SPECIMEN

13 (FIGURE 20).

34
FIGURE 22.
1nn
DETAIL OF THE EDGE CROSS SECTION OF SPECIMEN
13 (FIGURE 20).

- oon /4

FIGURE 23. TOP VIEW OF FORGEABILITY SPECIMEN 6A.

35
hc

im

FIGURE 24. SCHEMATIC ILLUSTRATION OF INHOMOGENOUS

DEFORMATION OF A CYLINDER DURING UPSETTING.
AFTER REFERENCE 47.

36

.1
't. .....
 material in contact with the forging die and the die itself, requiring lateral
flow of metal near the die to proceed by shear alone without sliding. A top
view of a buckled laminate cylinder is shown in Figure 23 in order to illus-
trate this frictional state. The lack of lateral material flow may be noted
in this figure. In addition, the cross sectional morphology of the transverse
forging shown in Figure 22 is a result of die friction and the anisotropy of
lateral flow. During upsetting large tensile stresses are generated in the
radial direction of the cylinder as a result of barrelling or non-homogeneous
deformation. The barrelling itself is a consequence of the difficulty in
flowing that the material adjacent to the die experiences as a result of fric-
tion, which, of course, inhibits outward flow directly, and local cooling of
the specimen through conduction to the colder dies, which increases the flow
stress of the cooler material. The consequences of these effects on the over-
all material flow in upsetting are illustrated schematically in Figure 24.
Region I in this figure, the so-called "dead metal zone", remains nearly unde-
formed while most of the deformation occurs in Region II,which pushes outward
deforming the material in Region III inhomogeneously. The morphological con-
sequence of this flow is shown quite markedly in the transverse specimen cross
section shown in Figure 20.
The phenomenon of hot or plastic buckling of the laminates observed for
the laminates must be treated somewhat differently from that of barrelling.
If each layer of the laminate cylinder is considered to be independent then
the buckling of the plates is governed by a stability equation similar to
Euler's column equation (pinned ends), 4 8
w2EI
Pcr = 'rE

where
1cr . the critical load for buckling,
E = the modulus of elasticity,
-. I = the moment of Inertia, and

L - the column length (height),

in which the modulus, E, is replaced by a plastic modulus, Et, that repre-
sents the slope of the stress strain curve in the plastic region at the stress
value of interest. 4 9 The magnitude of Et represents the work hardening
capacity of the material, i.e., Et for a perfectly plastic material would
equal zero and would have no resistance to buckling after the yield strength
had been exceeded. The moment of inertia for the lamina may be approximated by
37
 bh3

where
b - the plate (lamina) width and
h - the plate (lamina) thickness.
The critical load for buckling, of the lamina is, therefore,

P =2 Etbh3

SCr 12L 2

This equation then shows that the critical load is directly proportional to
the plate width, the cube of thickness, and the plastic modulus or work hard-
ening and inversely proportional to the square of the length or height. Since
the interlaminar strength of these laminates is not zero, however, there must
be an additional load required to separate them and the actual critical load
must exceed Pcr. Of course, if delamination did not obtain, then the lami-
nate cylinder would be expected to barrel and deform continously as do the
monolithic cylinders. The fact that this does not happen is a result of die
friction and the geometry of the individual laminae as explained below.
As noted previously the frictional condition for all the cylinder upset
testing was one of essentially pure sticking friction and this allowed the
modeling of the individual laminae as plates with pinned ends for the purpose
of estimating buckling strength of each lamina. Since the laminae are not of
equal width, however, their moments of inertia are not equal, the outer lami-
nae being considerably less wide than the central layers (Figure 6b). In
fact, the moment decreases as the chord of a circular section of the cylinder
or as
/
2 - d2 )1 2
b = 2(R

where
R = the cylinder radius and
d = the length of a normal from the layers to the center axis of the
cylinder.
The buckling strength of the outer layers is, therefore, less than the central
layers, and this leads to an inherent instability in the cylinders and favors
delamination as the outer layers tend to buckle earlier and more than the

38

#
 inner. This decrease in required buckling pressure is compensated partially,
however, by the increase in pressure due to the die friction itself. Never-
theless, this pressure increases toward the center of the cylinder as 1/d and
therefore, if sufficient pressure obtains at the outer layer for buckling,
there will be adequate pressure on the inner as well, although along the plane
perpendicular to the layers and passing through the cylinder axis the buckling
is nearly stable, that is, delamination does not occur (Figure 25). Having
established the instability of the laminate layers to buckling once they are
decoupled, it is necessary to consider the development of cracking that leads
to the decoupling and the actual buckling morphologies observed.
Delamination during upsetting is initiated primarily by the tensile
stresses developed during the inhomuogeneous deformation of the cylinders. As
noted this inhomogeneity is itself a result of friction and local cooling of
the forgings. The barrelling shown schematically in Figure 20 leads to very
high tensile hoop stresses in Region III, Indicated graphically by the diver-
gence of the scribe lines as deformation progresses. These tensile hoop
stresses lead to the onset of delamination in the laminate cylinders and to
edge cracking in monolithic forgings. In the case of monolithic material,
however, further cracking generally must be driven by the hoop stresses alone,
whereas the laminate cylinders become unstable with respect to the alternative
failure mode, buckling, as demonstrated above. It may be noted, in addition,
that along the plane perpendicular to the layers and passing through the
cylinder axis the buckling is almost stabilized and delamination does not
occur between every layer as shown in Figure 25. This may also be explained
by reference to Figure 20 since the radial stresses in Region II are essen-
tially compressive and do not favor delamination. Furthermore, the dead metal
zone, Region I, tends to shorten the effective or deforming lengths of the
plates in the central region and increases the local buckling strength. It is
Illustrative that the onset of central delamination shown in the specimen of
-. Figure 25 is at essentially the boundary of Regions II and III and that the
layers within Region III have buckled almost stably, i.e., together, since the
hoop stress is essentially in plane with them in this section and the plates
have similar moments of inertia. The dead-metal zone may also be perceived in
cross section in Figure 25 and compared with the schematic in Figure 20. This
final comparison can Illustrate the explanation for the partially stable upset
morphologies exhibited by some specimens and illustrated in Figure 19. In

39

- l: , . . t" °,.
1; I L-"
 4-i

FIGURE 25. CENTER PLANE SECTION OF SPECIMEN UP5B.

MACROETCHED IN NITRIC ACID. MAGNIFICATION: IX.

FIGURE 26. SIDE VIEW OF TRANSVERSE FORGING UP1O FRACTURED

TENSILE SPECIMEN SHOWING DELAMINATION AND
INDEPENDENT LAYER FRACTURE.

40

_ _ _
these specimens the dead metal zone sufficiently shortened the effective
length of the central layers that they upset without gross buckling or com-
pletely delaminating. Nevertheless, sufficient local buckling and delamina-
tion did obtain in all cases to preclude the relevant tensile or fracture
testing of this material.

3.2.1 Tensile Properties

The tensile properties of the base 30UM plate and the base 4145 billet are
listed in Tables 5 and 6, respectively, and the tensile properties of the
30014/1010 laminate as roll-bonded are listed in Table 7. Two comparisons are
of primary note with respect to the 30014 systems. First, the average
strength, both tensile and ultimate, of the 3UO/1010 laminate is less than
the monolithic 300M. This is due to the fraction of lower strength interleaf
material in the laminate. Since the volume fraction of 101U is approximately
9.5 percent, the decrease in ultimate strength may be explained as a volume
fraction effect by the rule of mixtures considering the tensile strength of
the 1010 to be approximately 40 ksi (276 MPa). The decrease in yield
strength, however, is greater than would be predicted by the rule of mixtures
and appears to be a result of the plastic constraint imposed on the 300M4 lay-
ers following the yield of the 101U inteleaves. This leads to a biaxial
tension-compression stress state In the layer or stronger material that re-
duces the apparent uniaxial yield strength. 50 Second, the average uniform
elongation and true strain at necking of the laminate is slightly but consis-
tently greater than the monolithic 300M. It is believed that this results
from the stabilizing influence of multiple layers on flow, i.e., an incipient
neck in one layer may be retarded from growing by the neighboring layers.
This observation will be discussed further subsequently with respect to forged
material and the interfacial strength of laminates.
As noted previously, 35 , 3 6 the advantageous tensile and fracture proper-
ties of metal-metal laminates are a result of the independent behavior of the
layers during fast fracture. This independence comes about through delamina-
tion of the layers because of the transverse tensile stresses developed by
necking in the tensile test and plastic contraction in fracture toughness
testing. In tension the laminate behaves elastically very much like the mono-
lithic material and even after the weaker interleaves have yielded the con-
traction is insufficient to cause delamination up to the onset of necking. In
41
 LL M ~ 00I

c
CL.J %J (

LOL) V4U C4 (VC

LLJb

Lii co 4d

m~ N

~Lii
-j

= 0)

o- I.- Z - 90 O

Lii
~-J CM
;m
E
r-
CA
Qi
00 0
CL L ~ - - -

-2C W -i - - -

-JJ
LJ ~ LA-
LA
t
-~
qiC
04
-

= cn - - -D

*u
V)
- r=Lfl
~ ~f) C
zi z>L.E

US' ZS,

a- -j NO
L^ c 42
 Ln OD*~j c

U.S qt CV L -
w. 04 Cj Lfl

En C.. - ( 40

K-(en (%i 4w a

o CD ~

zJ

LLi

-0
oL AZI
go
4D4
-* -

U.S LJ L L 0
-l kD-In t
LI

-JJ
- L J-. I-en ( -Lo0

CJ
m C'j C~j e..s

LUn

US

(A a - 43
 V.) to
Li = c 0% CJ co r
ce ~ 1 '4mLn
10 00 Le) r

tA u xA CJ . A 0
LA 0i -e LA c~
(A (%J .j C%j % CJ

VI1 C) co ko LA co qc

LaA
oL . 4 C 4 C
0I
C3-
0%u '0 0% L
C3 'i c

LD

co 0 -

A) - A - -
cc -L
2!
0C

C- Cn La- - -

LaL

L>

uj 0 a% m - cC "

btJ La 2: 0 - -r w %
~a-= Lo ko %D k0 LA

~I- ) 0% 0 0i

RE L~.b

0CD 0 0 0

v) -le C'j CV(~ CV

V) le c C DLA L C> C

Lao ~m Kr
A a'.
- cV
L ; * - 'ON CD

0
CVg.e n 0
00 0n

z v en. Lj
- -~

2C aa a w

Ii j -cc 's O'S

~ I I r-4 c

44
 fact, as noted above, the uniform elongation of the laminate is greater than
the monolith. Following necking, however, the transverse stresses lead to de-
lamination as shown in Figures 26-28 for a transversely forged specimen. The
tensile data for the transversely forged specimens are listed in Table 8. The
transverse forging of the as roll-bonded laminate did not significantly affect
strength, although uniform elongation was improved significantly and tensile
elongation was improved somewhat. This may be attributed to an improvement in
the interlaminar bond integrity resulting from the increased transverse flow
and compressive loading that obtains during forging. Following de'ir.iination
during gross specimen necking each layer independently necks and finally frac-
tures with each layer evincing a slant fracture. This independent behavior
also means that the actual or total reduction in area for the laminate ten-
siles is greater, although the tensile elongation may not be significantly
different. Finally, the strength of the forged laminate is indirectly related
to the forging strain. This was noted only with respect to the UP11 forging,
however, and resulted from the accidental forging of a section of the 1010
steel box used during roll bonding into the forging. There was, therefore, a
large volume fraction of lower strength material in this specimen. The sub-
ject of post-forged tensile properties will be discussed in greater detail
subsequently after the presentation of the closed die or complex flow forging
data.

3.2.2 tlicrography

The microstructure of the transverse forgings in the quenched and tempered

condition was not significantly altered versus the as roll-bonded laminate as
shown in Figure 29. In general, the density of small voids at the bond line
decreases after forging, compare Figure 13, and the spheroidite free zone dis-
appears, of course, in the heat treated condition. It may be noted that the
*. lighter gray fine tempered mar+ensite does penetrate beyond the small voids
and into the interleaf. This is probably a result of the diffusion of alloy-
ing elements into the low carbon interleaf, thereby improving its hardenabil-
ity. The microstructure of the interleaf is mixed pearlitic, bainitic, and
martensit ic.

45
 .-I I II I

FIGURE 27. DETAIL OF THE SPECIMEN OF FIGURE 26.

'I

FIGURE 28. TOP VIEW OF THE FRACTURE SURFACES OF A TENSILE

SPECIMEN MACHINED FROM TRANSVERSE FORGING UPlO.

46
 LU

zj a

Id 1) - 0% p- 0

uJ

cy-4

LL- =

C1-UJ
0D

LLI
I-

-j

0-4 ve CD
-M C" % CI
o U
L.-J

ol oUl
I..

Id) LI LI) 0 r
L)UJ C%i C14 C1 CI
LL.
I-
Idn -- 0% M~
_ Le)

0- %0 c

I.--
LU .-j C W) LI

LUI

a 00 L

LUU d. 0 0 I) I

aj C' 0 D.

4- 47
 100 lim

FIGURE 29. MICROGRAPH OF TRANSVERSE FORlNG UPIO, TRANSVERSE

SECTION. THE 1010 INTERLEAF IS AT THE LEFT.
MAGNIFICATION: 200X. 2% NITAL ETCHANT.

48
 3.2.3 Fracture Properties

The fracture toughness data for the baseline 30014 plate and 4145 billet
are listed in Tables 9 and 10, respectively. The important values for subse-
quent comparison are the average toughnesses and strength ratios. It is also
worttW of note that the monolithic toughness of the 3U014 plate is generally
slightly greater in the LT orientation. The highest toughness orientation for
the 4145 billet is the TS orientation. Since the toughness versus thickness
relationship was of importance for the 30014, several thicknesses were tested,
all of the LT orientation. For comparison and baseline purposes the fracture
toughness of the as roll-bonded 30014/1010 laminate was detemined and these
results are listed in Table 11. Similar comparison results may also be found
in Reference 35. Although the plane strain average or conditional fracture
toughness of the laminate is somewhat greater than equivalent monolithic
material, the significant effect of lamination is best seen through comparison
of the critical fracture toughness, an elastic-plastic toughness, that mea-
sures the resistance of the material to fast fracture or critical crack growth
more effectively than the conditional toughness. Since the Individual layers
in the laminate would not meet ASTM E 399 37 criteria for KIc, the tem
conditional fracture toughness is used for both laminate and monolithic mater-
ial. It may also be noted that the laminate strength ratios are much higher
than the monolithic alley. This increase in toughness in the laminate comes
about because the individual layers decouple near the onset of unstable crack
propagation and in thick laminates the stress state changes from one of essen-
tially pure plane strain closer to one of plane stress. This individual layer
behavior may be seen in the compact tension specimen shown in Figure 3U. Note
that the fatigue precrack fracture surface is flat and that shear lips are not
developed until fast fracture commences. In fact, it has been fractographi-
cally detemined in previous work by the author 36 that the delamination does
not occur until there is sufficient plastic contraction, that is within the
stretch zone. As a result of this decoupling the required critical fracture
toughness or energy for fast fracture is greatly enhanced.
The effect of forging on the fracture toughness of both the 300H/101U lam-
inate and the 4145 cylinders may be evaluated through reference to Tables 12
and 13. The toughness results for the 4145 forgings that are listed in Table
12 are somewhat higher in general than those for the 4145 billet (Table 10).

49

LL-
 0 0

= C i - ..

~LLJ
j CS
LO- (7%
en
0
%T
CC
en to
t
3 L
%0 mr
a.a, . Ll Y
*~ ~ ~ ~ ~~V (A tJc-zJ c, 0 n C,

LL'~ C co r. m3 to m 0

cr 40 &A 2: C%4 C6
0z - t*) 0
(JoL. U~-' 00 r 0 3 to -

"; C',J ;= -- - - -
C. 0jr 0 - -. 0 i' 3 L
LA C 03 ' r-, co o r- c c
O c w g I3 O C'J LA 0D 1. 1 CO
a', o.; 06 CJ C;
U- nxC41
CL e
dr- tD r- W) ,P- cm Im tD
LiD

UJ c3L S 00 0* 0 4Z 0 0 0 0 0-

C) ClC) 00toL) to * C> ('4

r% m0
V). 4c .- % 03 " . L
V) LL 2 N r - ko - n - C'
00 C'o
Z u'i
C-0' - 0 A A ~ t
C) ~ C .' ~
Loe-cJ U- A L 9' - t

U.)
j2 un c

*r S I It!I
S C1 *

L S I I S

I-V)- - ) I - I

5--

a~ ar a a *AL

Ixw 0 0 -j Zo-

rn f" M) C -cc i z * z-
 ui
~c
=0 ac Co cn
* C%j qr ff)
L) LLJ 0
a- m~ rn mr ko
LLJ wC~j (%j ('j (%J m~ ej cl-i

_j UA V). L (.

~LAJ - m LA in 'c en e ~
- CVF -~ *- - ,- - - -

uj~
:, (A

-j ~ ~ (NJ r n-Mc c n q w

I -) 1 I ON m
co a- -c *. I'-_
-n cr U- en
-3- 1- V) q

aU0 AZ- c Lj . 0 L
CD L IO - -% - - -Y -v -~

0n CV m-o c %
V) C -d ' 0J cc
0; C6
L C-

le CD) 0 0i; _ LA (N.

0, ;c
C3 i- Ad
*ri -* -:

ci OLj w c cc A - - c

I.-

L" u(N - - - -A
co -- az Ln a" ko Lmc
< Li A1 co (J C t ~
cc
Ul P-
L Ln LA Ln

l.LI
~D.1iE
zE

- ~ ~ u >J-F I I;

-- J

<4 q..-L If; LI I Ll

t; V) le
*.LI' K a a ca e ar a a

4 .8 a .8 . .851
 LLJ

L;c& 0~
co Go M~ r- -4
U LUCD YD. (n r-. r-.
LLJ ~ oe 00 P%
0- 0

-j LIi A .- - -
40 M- ON Ln CV)

C3 cc W cc ul %r~ w r-

ex-)J-Jk
0WU -W C) C

LA CA-< M -
M-i~.
= to0
-- C'J
La fpc CY%0
-I.

(/I 0 e0o 0 en 0

LUn

- -

- 11L cc Cl a
0l a

?- ifal,- L

0.

m% 0% CD

iu
uJ
0.
U- *- c L
. *x

Ln

~ oc ~ 52
 I I

(b)

FIGURE 30. 300M/I11 LAM4INATE COMPACT TENSION SPECIMEN

MM3 FRACTURE SURFACES. (a) TOP VIEW,
(b) END VIEW.

53

- .
 '.AJ L

0 0 0-

v CJC ~ % _

I= L
cc J 0; cs ci(l

CI en LO %0

CD I-LJL

;c x L LA; r-: 1- (1

LL-
o i
ce
2c
=
q
-
Oc
-
Ln UlA

ccI uii .0 ~ ~ (n m
a. 0

LO

~ OLLL. 'C N *U

o i:~ rE LA Ln

V) OCJC.0 I fa oi CO ' 0

U.Jr- 0t0 Ul) 0 c5

* C)

tu I Ll
=)
LLJi
Vc-a
<'.
-
%
c

-l
LA
r

-
CO
r

LAJ

A .0
UJ I- J *

1- LA C

LA
'm cia

La IA- ;.i;i-

~
-J J U, 54
 LjJI
CoLJ- 00 0
mA '%
uj CZ 4S .- C% CJ C'J C%

aC%J__
CIJ +
LU CA s-r
~UJ
Cz *LO 0CD r r-.. CD
LiiIJ- C%j (%j 04. COj C~j C14
In t- LO to ~- c - - - -

:w (-)LL - - C~J I C'J C~'j % (J

V)

Cl W Lii =n - I - -_
U- ZL) *M to a, CJ

al. * = CL ON I m) 0% fl- co
LI) Cl- Ifl 0D A 04 CD
au, C% c J cJ
uj- U

to - - Zrf-
o0 - - .Co mO-.-
(AL %0 0%oC.
Co 5%. 0-be

2LA 0~ NA I,,-
.n

0i Q- U' 0% a -

jfi cm
o VJ C' C\J LA 0! O.

!0
00 P* f* a, a, a, -

ko - T -

V; LO C% L 0
Ii- ( A) C" , aja C-i. L
(u) -!) =- LA 10 0 --

*j 01 0 0l 0 0 0)-

o~L 0

I- I-I- ~ .J - -as- -

z w

a. a. 0. C3. a. C3. p LAa

49

55
This Improvement is a result of the additional flow or fiber developed during
upsetting, and both of the specimen orientations tested, i.e., LT and SL, are
of the high toughness "cross grain" type. Some improvement in toughness in
the LT orientation may also be noted with increasing upset strain as the flow
and fiber in this orientation is further refined. There was also a slight
difference between the top and mid-plane toughness in forging 7A in the LT
orientation, although this is not considered significant and probably arises
from slight orientational differences with respect to the fiber of the upset
cylinder.
The fiber of metal-metal laminates on the other hand Is Intrinsically de-
termined by the arrangement of the laminae. Since laminate transverse forg-
ings flow in the two directions of the lamellar planes the LT orientation re-
mains rigorously "cross grain" or cross fiber in these forgings as it was in
the parent laminate plate. The consistently improved toughness of the trans-
verse laminate forgings may be appreciated by comparing Table 13 with Table
11. Since the fiber orientation between all these specimens is the same, how-
ever, the improved toughness must result from flow or other geometric differ-
ences. In fact, the primary improvement in toughness is a result of the thin-
ner layers in the forged laminate, which shifts the decoupled stress state
further toward the tougher plane stress condition, note Figure 4. The subject
of layer thickness versus toughness will be discussed in greater detail subse-
quently.

3.3 COMPLEX FLOW EVALUATION (CLOSED DIE FORGING)

The metal laminate cylinders forged in a closed die to form the part shown
in Figure 7 demonstrated excellent forgeability in this complex flow geometry
both with respect to the stability of the layer versus interleaf flow and the
ability to flow in three dimensions, that is perpendicular to the lawellar
plane. In addition, the intrinsic fiber of the laminate acted as an extremely
effective tracer for the flow patterns developed during the closed die forging
of the part geometry investigated. The as forged appearance of the first
closed die forging (Cii) is shown in Figure 31. The uneven edge material in
this forging is flash and some delamination of the laminate did obtain in the
flash as may be noted in the figure. Nevertheless, the material flowed very
well and stably as shown in the cross section of Cl1 reproduced in Figure 32.

56
 FIGURE 31. CLOSED DIE FORGING CD1.
TOP VIEW.

FIGURE 32. RADIAL CROSS SECTION OF CLOSED DIE FORGING CD1.

MACROETCHED IN NITRIC ACID.

57
 The flow is illustrated very well in this figure including the reversal of
flow that occurs in the outer flange as the flowing steel reaches the flash
gutter of the die. The delamination that developed between the sixth and
seventh layers from the top extended approximately one inch into the forging
at this point. This delamination developed as the material began to flow into
the flange section as shown in Figures 33 and 34 of forging C02. This forging
did not fill the die but rather stopped flowing just after the material had
reached the flange cavity and begun to spread. The tensile stresses developed
as a result of die friction as the forging began to flow into the flange cav-
ity were sufficient to cause delamination. Although these delaininations may
shut during subsequent die filling they will usually result in seams as shown
in Figure 32. Whenever the stress state is basically compressive, however, as
it is in the hub cavity of the die, the dual metal laminate flows stably with-
out separation of layers and interleaves. This may be seen quite well in the
etched cross sections shown in Figures 32 and 34. The closed die forging,
complex flow evaluation may be considered to give rise to a flow and stress
state intermediate between that encountered in the longitudinal and transverse
open die forging. The successful forging of laminates in a closed die are
contingent upon proper die design in which flow, particularly divergent flow
such as that into the flange cavity, is prevented or inhibited through the use
of generous billet radii and draft.

3.3.1 Micrography

Representative microstructures from the web area of a closed die forging

are shown in Figures 35-37. In general, the flow obtaining during forging ap-
pears to improve the integrity of the layer-interleaf bond as notable in Fig-
ure 37. The excellent stability of the flow is also markedly illustrated in
the low magnification micrograph of Figure 35. The flow stability of lami-
nates is the second most important criterion for the successful forging of
these materials, the primary one being a basically compressive stress state
perpendicular to the lamellar plane. Even at the very high strains, -107 per-
cent to -154 percent, produced during the closed die forging, however, the
layer and interleaf materials deformed almost uniformly and stably. Such
stable deformation appears to be primarily contingent upon the existence of

V ____
58
 FIGURE 33. THREE QUARTERS VIEW OF CLOSED DIE FORGING CD2
AS FORGED.

LL f ? 1t

FIGURE 34. RADIAL CROSS SECTION OF CLOSED DIE FORGING CD2.

MACROETCHED IN NITRIC ACID.

59
 .. . ., , ,.........

0.5 pm

FIGURE 35. MICROGRAPH OF CLOSED DIE FORGING CD3, TRANSVERSE

SECTION. MAGNIFICATION: 44X. 2% NITAL ETCHANT.

FIGURE 36. DETAIL OF FIGURE 35. MAGNIFICATION: 500X.

2% NITAL ETCHANT.

60

_ _ _ _ _ _-----------
 tA

. 50 pm

FIGURE 37. DETAIL OF FIGURE 35. THE 1010 INTERLEAF IS AT

THE RIGHT. MAGNIFICATION: 500X. 2% NITAL
ETCHANT.

61
 approximately equal flow stresses in the two materials at the operational de-
formation temperature. Previous unpublished work by the author on the hot de-
formation of the beta titanium alloys suggests that the material with the
higher flow stress will tend to breakup or fragment during deformation. This
is also in agreement with earlier work by Semiatin and Piehler 5 l on the rooa
temperature deformation of clad sheet materials in rolling.
The microstructural transformation products of the closed die forgings are
similar to the other heat treated laminates with the exception of less pro-
eutectoid product appearing in the 1010. Since these structures are represen-
tative of the relatively thin web, the 1010 interleaf is nearly coiapletely
transformed to a bainite, while the 300 is temperated hlartensite as expec-
ted. The line of demarcation between the martensitic and bainitic structures
is once again quite clear, although the few lighter etching patches in the
300M near this line may be evidence of the original bondline as discussed in
Section 3.1.1. The relatively larger prior austenite grain size and wide
grain size variation is probably a remnant of the rather high, 240J0 F
(1316 0 C), forging temperature used.

3.3.2 Tensile Properties

The tensile properties of the several closed die forgings tested are
listed in Table 14. A comparison of these results with Tables 7 and 8 will
reveal that, in general, both the tensile strength and yield strength of the
closed die laminates is lower than the as roll-bonded or transversely forged
laminate. This effect is illustrated in Figure 38 in which the strength is
plotted versus strain for the various material systems. The tensile strength
of the closed die forging is, in fact, less than that predicted by the rule of
mixtures and suggests either a difference in the intrinsic strength of the
300H layers in these forgings or a change based on the structural differences,
such as the thinner layers and interleaves and the improved bond. It is felt
that the latter, structural differences, are probably more instrumental in
this regard for the following reasons. First, although the volume fraction is
approximately the same, the thinner layers are more effectively stressed,
i.e., a greater fraction of each layer is affected, by the plastic contraction
of the interleaf as discussed in Section 3.2.1. Second, as a result of the
* improved bond integrity this traction is continued to higher levels of tensile

I, 62
 LLfa
2C ma cc cl 0 ma

IL Un (% % .r % ' C%J

CZ

LU-
-n I c n) c c
0A I I en*j *n m -t i

-
LL)

cm a:-C

LLJI
V)(I I ~ L % n L ~ ~

L zg I ON t co m% r= 0% co

LUI fa a% -S - .- - - -

n L %0 0
C) %D 0)C
< -j C. x r. 'J ai C) cl LA Cla Li L
LA *-- - - -~ -l
Ln %D

-iLhiU
_j o C~j !. en, in ko m~ in C
Lii en C',
mA m M m~ en enj
-~ -l - '
Cs(J C'J e4j CJ (14

LI..
m. Li.. 0%
4Di C Z - - - - - - -
Il LA r- co - CD 00 -
- A CL LO. 0 - La 0 (V) M
0A.s~ La 0%) 3%0 0 %
rn %

z C0-in. cW
00 # - - -
Lii

-i LJb 0 0 LAn n LA LA LA LA LA
Lki CL .- LU

(-~A C..

5- La 3% 10 0 '0 a '0 La co
f., f-
q')-
Ln
-..
'I
m .
0
en
-
0~
n
~ i

LU 0 0 0 0 0 0

'Or 4' C, M, I

-
CL fX
'.A m -
m5-

a 5-63
I-.-'

Lii LiJ~=0'L
 t 1I 1 I I I

0 300M MONOLITH 2.0

280 jJ 300M/1010 LAMINATE,

ROLL-BONDED 1.9
A 300M/1OO LAMINATE,
* ROLL-BONDED AND
260 FORGED TRANSVERSELY - 1.8
V 300M/1OlO LAMINATE,
R B AND FORGED
ANDROLL-BONDED
IN A CLOSED DIE - 1.7

240
0 1.6

, 220 - 1.5 "'

- 1.4
200 -

- 1.3

180 -
OPEN POINTS: 0.2% YIELD STRENGTH 1.2
CLOSED POINTS: ULTIMATE TENSILE STRENGTH

160t 1

4. 0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8

TRUE UPSET STRAIN

FIGURE 38. YIELD AND TENSILE STRENGTH VERSUS UPSET STRAIN FOR THE 300M
MONOLITHIC AND 30014/1010 LAMINATE SYSTEMS.

64

...
. . .. .----- 7- .. .j...
..
 elongation, thereby further depressing the ultimate tensile strength for all
forged material versus the as roll-bonded laminate. That this greater elonga-
tion did obtain for the forgings is demonstrated by the data of Figure 39 in
which elongation is plotted versus upset strain. In this figure the trend to-
ward improved uniform elongation in laminates noted previously 3 5 ' 3 6 is de-
monstrated very well. Although it is somewhat speculative at this point, it
is felt that an optimum layer-interleaf bond strength may exist for the best
combination of strength and tensile ductility or formability, since uniform
52
elongation may be correlated well with formability in many alloy systems.
Previous tensile results repoted by Johnson 3 3 for a Ti-6AI-4V/CP titanium
laminate in which delamination did not occur suggest to the current author
that it is possible to obtain a system for which the bond strength is too
great for individual layer behavior either in tension or in sharp crack frac-
ture testing. This, then, leads to laminate behavior that is indistinguish-
able from the monolithic. The consequences of this more monolithic behavior
are reflected in the decreasing total elongation with improved bond strength
as displayed in Figure 39. Nevertheless, the laminae did behave independently
in tensile fracture even in the most severely deformed closed die forging.
This is illustrated in Figures 40 and 41. Once again the actual reduction in
area of the laminate is much greater than a corresponding monolithic ,waterial
since each layer necks individually. It may also be noted in Figures 40 and
41 that as the layer thickness is decreased the individua, layers begin to be-
I have more as if they were thin sheets, i.e., their tensile elongation is trun-
cated vis "a vis thick specimens following necking, 53 and this may also con-
tribute to the decreased total elongation of the laminates with increased up-
set strain. Finally, because of the different themal-mechanical history of
the closed die forgings, especially the high forging temperature, some varia-
bility in their tensile properties may be an effect of metallurjical condi-
tion, although this effect is almost certainly very secondary to the conse-
quences of the structural differences.

3.3.3 Fracture Properties

The fracture toughness properties of the closed die forgings listed in

Table 15 are in general agreement with the statement that the toughness, es-
pecially the critical fracture toughness, improves with decreasing layer

65

J1
 14 0 300M MONOLITH
jj 3001/1010 LAMINATE
13- ROLL-BONDED
A 300M/1010 LAMINATE
ROLL-BONDED AND FORGED
12 TRANSVERSELY
7 30014/1010 LAMINATE
ROLL-BONDED AND FORGED
IN A CLOSED DIE

10

8
i g 7-
I-

, z 6
.-J
Li.i

2 OPEN POINTS: TRUE UNIFORM ELONGATION

CLOSED POINTS: TOTAL ELONGATION
1 -
I,-I
; I I i I i I I 1 i I
0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8

TRUE UPSET STRAIN

FIGURE 39. UNIFORM AND TOTAL TENSILE ELONGATION VERSUS UPSET STRAIN FOR
THE 300M MONOLITHIC AND 3001/1010 LAMINATE SYSTEMS.

66

7 .-
 FIGURE 40. SIDE VIEW OF A FRACTURED TENSILE
SPECIMEN MACHINED FROM CLOSED DIE FORGING
CD5.

FIGURE 41. VIEW OF THE MATING TENSILE FRACTURE SURFACES

OF CLOSED DIE FORGING CD5.

67

-
-
0-
.. =.,........- -
 thickness. Nevertheless, a comparison of Table 15 with Table 13 will reveal
that the transversely forged laminates were somewhat tougher across the spec-
trum than the closed die forgings. This is believed to result from the im-
proved bond strength of the closed die forged material. Such improved
strength would lead to a delay in the onset of delamination and the imposition
of a greater triaxial component during fast fracture giving rise to a more
plane strain condition and consequently lower toughness. Nevertheless, the
individual layers still do delaminate and fail in mixed mode as shown in Fig-
ure 42. In addition to the improved bond strength, the wavy or non-linear
character of the layers and their uneven thicknesses may have affected tough-
ness, although it is not felt that this would have been sufficient to have ac-
counted for a significant fraction of the observed difference in toughness.
Rather it is more probable that with regard to the toughness as well as the
tensile ductility (See previous section.) there is an optimum level of bond
strength.

3.4 GENERAL DISCUSSION OF LAMINATE FRACTURE TOUGHNESS

As pointed out previously above and in other publications 4,34,35,36 the

two primary criteria for improved fracture toughness in metal-metal laminates
are (1) a thin layer thickness in which plane strain conditions do not obtain
and (2) the mechanical decoupling of the layers during fast fracture. It has
also been demonstrated previously35 that when these conditions obtain with-
out qualification the thin layer toughness may be maintained in laminate
panels of indefinite thickness. This is demonstrated well by plotting the
critical toughness, an elastic-plastic parameter, for the 300M/1010 laminates
In all conditions versus the toughness of monolithic 300M of the same heat
treatment as has been done in Figure 43. With the exception of the one trans-
versely forged datum, representing a testing machine malfunction, the tough-
ness of the laminates is maintained at mixed mode, i.e., between plane strain
and plane stress levels even in panel thicknesses at which the usual plane
strain criteria would be expected to apply. The line plotted to represent the
laminate Kc is a simple least squares 54 fit to all the laminate data ex-
cluding the one outlying point. The line represents essentially no correla-
tion between toughness and panel thickness. Most of the scatter notable in
the laminate data of Figure 43 may be removed, in fact, by replotting the same

68
{ ~ -
 0> CA Ln C%j D 0D Lc)
UAJ~
CcON ONh Ch 0% at 0% at 0%

K. .1
V)
CD

I-j C.) CD
-~

c
~
r-

~~LJl:
w
LLJ
c

CD 02m
*-

a
0

__
CJ
-

--

0
) 0

C 00
C0

n
m
hL
0

o~
%0
O%

-
0
O
M
Ch

0
-
0
0

V;JC cm0n %0 '. %a 0% qt qr

tA4 a: U fZL cJLA - O h O

L. a- < - -D % - -
Sj Cc
0- (A x
(D .~\ A~ A
0c LaiV co.
* .

om toZ~ v-.
Oh koc

C) w~ Ll U; -3 0l 0: LA C4 r-;
o I . ~ co m a cc o'

U-

V) f , -.. qw -W C~l C~j co cc

V) JuiJ 0 0 LOl LA LO Ln

-. 1! 7 * . . .I

V) LA .- 0 0 LA

* LA qLA J C CJ (%D evn 0

C) C; I-.A N CA LA li
LA -D Qh tC) C> a~-

w iC 0
ClClC C
-; E

a- l.a M * cm I-~
C) Co*D 0 al *D -Co

a3ac a & a.
wa * a z le

69
IT lo
w-

FIGURE 42. COMPACT TENSION SPECIMEN FRACTURE SURFACES

OF CLOSED DIE FORGING CD5.

I 70
 W-edW bSS3N19~i 3IW JIA VII

wo -W N~ 0 %D qr C~ 0
Cl to qw o

00
CO -

o; C

LALA- 0
* c)
V V) Cl
>. L.)

0 0 0 0 0 0
N v1 < LA.'

- r -- *4~~:< 1

C3 1
critical toughness data versus layer thickness, the layer thickness of the
monolithic material being tantamount to the total panel thickness, of course.
This has been done in Figure 44. In general, the toughness of the laminates
lies along the monolithic toughness versus thickness curve, the divergences
being rationalized in tems of the bond strength and volume of low strength
interleaf as discussed in Sections 3.2.3 and 3.3.3. Briefly, in review, the
as roll-bonded laminate will have a slightly greater toughness than the mono-
lithic 30014 because of its low bond strength, giving rise to nearly complete
layer independency, and to the energy absorbed by the low strength ductile
1010 interleaf. The forged laminates tend to have toughnesses slightly less
than the monolithic material primarly is a result of the improved bond
strength and somewhat more constrained mechanics, that is a shifting toward
plane strain. Such a shift also explains the lower thoughness of the closed
die versus the transverse forgings.
The effect of stress state and elastic constraint on the toughness of lam-
inates as compared with monolithic 300M may be deduced from the data of Figure
45 in which the conditional fracture toughness or resistance to subcritical
crack growth, a linear elastic parameter is plotted versus total panel thick-
ness. Since this fracture toughness parameter essentially represents the in-
ception of crack advance, the metal-metal laminate property of independent
layer behavior has not had the opportunity to develop and, therefore, the con-
ditional toughness is not greatly higher than the monolithic 30UH. The trend
lines are detenmined once again by the method of least squares. It is worthy
of note, however, that even the linear elastic toughness of the laminate is
never lower that the monolithic plate with the exception of the one outlying
point due to machine malfunction and representing essentially a dynamic frac-
ture toughness. Furthermore, the trend line for the laminate lies everywhere
above the monolithic material within the panel thickness range of plane strain
monolithic toughness.

3.5 GENERAL DISCUSSION OF MATERIAL FLOW AND FRACTURE TOUGHNESS

It has been known 5 5 generally for some time that the proper orientation
of grair flow or fiber in forgings improves mechanical integrity and toughness
as a function of design and loading geometry. In particular, toughness is Im-
proved in the "cross grain" orientation in which the crack must propagate

72
 AVERAGE LAYER THICKNESS, m

1 2 3 4 5 6 7 8 9 10 11 12 13 14
I I I I I I I I I I I I I 260

220 0 300M MONOLITH 240

200 A
& 0 300M/lOlO LAMINATE, ROLL-BONDED 220
III A 300M/010 LAMINATE, ROLL-BONDED
r 180 -- AND FORGED TRANSVERSELY - 200 0E

160 300M/1010 LAMINATE, ROLL-BONDED 18O

AND FORGED IN A CLOSED DIE
- 160
uj 140
- 140
C- 120
-120 Lu
S 100 --
100
U-
'- 80 - 80 -
60 - ~0
- 60
C
4O

40- 40

20 - 20
I I a I I , I i I
0.1 0.2 0.3 0.4 0.5
AVERAGE LAYER THICKNESS, in.

FIGURE 44. CRITICAL FRACTURE TOUGHNESS VERSUS LAYER THICKNESS FOR MONOLITHIC
300M AND THE ROLL-BONDED AND FORGED 300t/1010 LAMINATES. THE
NUMBERS WITHIN THE DATA REFER TO THE INDIVIDUAL TEST SPECIMENS:
SEE TABLES.

73

-- -- _ i _- -------
 0 0 0o 0D 0~ 00 0 0 0o 0i 0~
04 04
C~ 0c Ol 4 0

0D 4
-cc
CIQ0

'0m o

-U o U

Li Li Ln

L3I 0 V)J
LOA

.. L
c.*J L- 1cLc 0J C-

u L- UL- L

0l
C) 0M V) Vfl

Lii o4m Lii

2 0 M00
-i ~~~ 0
co oI L
+- Q

Lii 17 U7 Li CD

Q 04C.C
ZC m
i-i~~C .L <-

=v
CDC-CD C

Z/L~

- 0 0 04
 across the fiber. The effects on toughness of both fiber or orientation and
panel thickness in addition to the improvements obtainable through forging
flow are shown in Figure 46. In this figure the toughness of the 4145 alloy
steel used for forgeability comparison is plotted as a function of panel
thickness and specimen orientation. As may be noted in the figure the im-
proved fiber obtained through upset forging has improved the toughness vis a
vis the billet for all equivalent orientations. It also is worth noting that
for the 4145 both the conditional toughness, KQ, and the critical toughness,
Kc, are improved significantly by forging. Such an upsetting process for
i billet is a common practice within the forging industry prior to final foring
of more or less tabular parts in order to fold the longitudinal fiber of the

billet parallel to the forging axis for improved radial or longitudinal pro-
perties in the finished forging. The relatively high toughness of the TS
orientation in the 4145 billet may seem somewhat surprising at first until it
is remembered that this is a cross grain orientation and, in fact, a visual
examination of the fracture surface, shown in Figure 47,will reveal a very
coarse fiber representative of the billet center.
When the critical fracture toughness is plotted versus upset strain, a
measure of the extent of flow, as has been done in Figure 48, the monolithic
4145 exhibits generally increasing toughness with increasing strain. This is
attributable to the improved fiber of the monolithic material as discussed in
the preceeding paragraph. The laminate, however, with its intrinsic fiber
tends to peak in toughness at the intennediate strains associated with the
transversely forged laminates. The least squares trend line is still of posi-
tive slope; however, the behavior of the data may be better represented by the
concave downward curve that is sketched in as well. As noted in the previous
section this behavior is a consequence not of improved or worsened fiber, but
rather of the change in mechanical behavior associated with both the layer
thickness and the bond strength.

3.6 FATIGUE STRENGTH OF ROLL-BONDED LAMINATES

The results of the fatigue testing of as roll-bonded laminate plate and

30014 monolithic plate are plotted as a W5hler curve in Figure 49. It may be
noted that the fatigue strength of the laminate is higher, in general, over

75

* - - -- - "-r
 PANEL THICKNESS, mm
14 15 16 17 18 19 20
-- 7I 8
I I 10I 11I "121 13I
9 f I I I I I I

70

60
LT SL
60

lu 50 LT //

E J
40 - *- TS

30 00
LT3L 3

TLi

20 04145 BILLET, KQ 20
* 4145 BILLET, Kc

10 0 4145 UPSET FORGING, KQ

M 4145 UPSET FORGING, KC 10

0.3 0.4 0.5 0.6 0.7

PANEL THICKNESS, in.

FIGURE 46. FRACTURE TOUGHNESS VERSUS PANEL THICKNESS FOR THE

4145 BILLET AND UPSET FORGINGS.

76
I
- - - ,, . .. III ' ' - - - -.- :" -
 34'S !
FIGURE 47. COMPACT TENSION SPECIMEN FRACTURE SURFACES
OF 4145 BILLET, TS ORIENTATION.

77
 260

220 -240

0i. 200 -. ~- 220

00

- 160 ,' K 15.9c +174.5 A180

£m 160 L2"
CD 140 - (

I-* 300M MONOLITH -140

120 -U
300M/10lO LAMINATE 10~-
S 100
U- * ROLL-BONDED lo
10 u
S 80 E ROLL-BONDED AND c
FORGED TRANSVERSELY - 80
S 60 - A ROLL-BONDED AW~ FORGED 60 (0
0 0 IN A CLOSED DIE
40 TS 4145 40

20 -0 BILLET 20
o UPSET FORGED

0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8

TRUE UPSET STRAIN

FIGURE 48. CRITICAL FRACTURE TOUGHNESS VERSUS UPSET STRAIN FOR THE
4145 AND 300M4 MONOLITHIC MATERIALS AND THE 300(4/1010
LAMINATE.

4S

78
'2dW 'SS3U.S OII3AJ WflWIXVW
CD
0%0 o 0 0c 0N

- C)

C,,

C=0
L-4

I-
o 0CD
-L Sd,

o o l

cyLL cY , LI

LA cA

C)
-

-LJ

c-i

Lii*

LL.
CD~~~i
D a (D a D CD C

00 %D
D qtt %D t C~

79
 the entire range examined, although it is difficult to determine if any signi-
ficant improvement in the fatigue limit is obtained. Nevertheless, based on
the relatively few data collected there does appear to be a significant im-
provement in fatigue strength at the higher stress levels. This improvement
appears to stem primarily from the improved flaw tolerance of the laminate
plate, such that final fast fracture of the laminate fatigue specimens was de-
layed even after the nucleation and subcritical growth of a crack in one lay-
er. This flaw tolerance is primarily dependent on the crack arresting proper-
ties of the soft interleaf, although the improved fracture resistance in the
crack divider orientation may contribute as well.

3.7 STRESS-CORROSION CRACKING OF ROLL-BONDED LAMINATES

The results of the stress-corrosion cracking evaluation are presented in

graphical form in Figure 50. Two primary conclusions may be drawn from these
data as follows: (1) The crack propagation rate at high stress intensities
does not appear to differ significantly between the laminate and monolithic
specimens in the crack divider orientation and (2) The threshold or critical
stress intensity factor for stress-corrosion cracking, is significantly higher
for the laminate material. The actual lower measured values for stress-
corrosion cracking were 39.1, 33.6, and 32.3 ksi-in1 / 2 (43.0, 36.9, and 35.5
MPa-m I / 2)
for the laminates and 14.9 ksi-inl / 2 (16.4 MPa-m 112 )for the
monolithic 300M. These values may be taken conservatively to represent
K scc. The propagation rate and KIscc results for the 300U1 also agree
well with those reported in The Damage Tolerant Design Handbook 5 6 , which
lists a maximum rate of approximately 4.2 (10- 5 ) in./s (1.U6(1tr 6 ) m/s)
and a Klscc of 15 ksi-in. 1 /2 (17 MPa-m 1 / 2 ) for 300M of similar yield
strength and heat treatment and tested in a 3.5 perent sodium chloride aque-
ous solution.
- The critical stress intensity factor represents the upper bound of load
for a flawed material in a specified environment such that no subcritical
crack propagation is expected, i.e., the material is safe from stress corro-
sion crack advance. This in turn allows a usable strength to be specified for
an unprotected material in an environment of interest. The above noted re-
sults, then, permit the inference that the laminate has a usable unprotected
strength greater than twice that of monolithic 3004 in the test environrent.

0
- ---. - ... - - - =r
80
~ - - - - - - -
 STRESS INTENSITY, MPa-m

35 40 45
104 - 5I 10I I
15 20
I 25 I 30i I I

1010-

o I°
o1__.d----
E
° 10 5- --

7
CD

C-CD

06
o
CD
C= 10

cn

--10-8

10- 7-
5 10 15 20 25 30 35 40 45 50

STRESS INTENSITY, ksi -in.h

300M STEEL
o LT ORIENTATION, 3.5% NaCi, 740 F (23.3 0 C)
300M/1010 LAMINATE, CRACK DIVIDER ORIENTATION AVERAGE
0 F (23.3 0C) LAYER
0*34 LAYER,
LAYER, 3.5% NaC1,7474OF
3% NaCl, (23.3°4)LAE
S4 LAYER, 3.5% NaCI, 74OF (23.3oC) THICKNESS:
A4 LAYER, 3.5% NaC, 740 F 0.125 in. (3.18 mm)

FIGURE 50. STRESS INTENSITY VERSUS STRESS CORROSION CRACK GROWTH

RATE FOR 300M STEEL (235 ksi (1620 MPa) YIELD STRENGTH)
AND 300M/1010 LAMINATE (201 kst (1386 MPa) YIELD STRENGTH).
SELF-LOADED COMPACT TENSION SPECIMENS, LT ORIENTATION.
81

LJ.- .go.
 In a protected condition the laminate would be considered to have, at least, a
better intrinsic factor of safety with respect to stress corrosion crack ad-
vance. In addition, the improved toughness of these laminates would require a
longer flaw for catastrophic failure without regard to the cause of the flaw.
As mentioned above these results and conclusions pertain to the crack divider
orientation. It is also expected that the crack arresting properties of the
soft interleaf would be highly effective in retarding stress-corrosion crack
advance. Unfortunately, experimental difficulties encountered in precracking
the bolt loaded crack arrest orientation compact tension specimens precluded
the collection of stress-corrosion-cracking data in that orientation. At the
time of this discovery insufficient material was available for the fabrication
of bend or cantilever beam specimens for testing.
The efficacy of the metal-metal laminate system examined in improving the
stress-corrosion cracking resistance of an ultrahigh strength alloy steel is
demonstrated very effectively by the data of Figure 50. The important deter-
minants of this improved behavior, however, are incompletely known or incom-
pletely understood. Several observations of the stress-corrosion fracture
surfaces can help at least to rationalize the cracking behavior of the laai-
nate. First, as may be noted in Figure 51 the stress-corrosion crack front,
which is approximately one-half inch from the end (left side) of the specimen,
is convex within each individual layer indicative of crack pinning or retarda-
tion by the interleaf or layer-interleaf interface. Second, although it is
difficult to distinguish reflectivities in Figure 51, the 101U interleaf frac-
tures are shiny and not dull and covered with corrosion product as are the
30014 layer stress-corrosion fracture surfaces. From these observations it is
inferred that in the crack divider orientation the interleaf and its attendant
interfaces acts to pin the advancing crack, bowing it and slowing or prevent-
ing its progress, by primarily a mechanical mechanism. Since the 101U is not
as susceptible to stress-corrosion cracking as the 3004, it pins the crack un-
til sufficient tunneling occurs and the 1010 fails by tearing. This tunneling
also effectively blunts and broadens the stress-corrosion crack front. Fur-
thermore, there appears to be an electrochemical interaction that occurs be-
tween the 1010 ijiterleaf and 30014 layer, and, based on the appearance of the
two materials, the 1010 is cathodic. Although the crack dividing and primary
action of the interleaves may provide some increase in the critical stress in-
tensity through purely mechanical means, it is difficult to rationalize the

82
 FIGURE 51. FRACTUE SURFACES OF 300M/1OlO LAMINATE
STRESS-CORROSION SPECIMEN TESTED IN AOUEOUS
3.5% SODIUM CHLORIDE SOLUTION AT ROOM
TEMPERATURE.

Ie

1 mm

FIGURE 52. OPTICAL FRACTOGRAPH OF A STRESS-CORROSION

CRACKING (SCC) SPECIMEN SHOWING THE ARRESTED
STRESS-CORROSION CRACK ON THE LEFT AND FAST
FRACTURE SURFACE ON THE RIGHT.
MAGNIFICATION: 34X.
83
improvement in mechanical terms alone. An electrochemical explanation can be
advanced, however, that includes both a mechanical and chemical aspect and is
based primarily only on the cathodic relationship of the interleaf to the lay-
er. Figure 52 reproduces a photograph of the arrested stress corrosion speci-
men and may assist in the proposed explanation. The limit of crack advance is
marked by the relatively flat region at the left of the fractography. Shear
lips are developed during fast fracture when the specimen was sepearated after
cooling to the temperature of liquid nitrogen in order to expose the fracture
surfaces for observation. These shear lips begin essentially at the point of
stress corrosion crack advance. Since the 3UOH layer crack tip is anodic, it
should be loosing material at a rate much faster than that usually associated
with stress-corrosion cracking. It is believed that this leads to significant
blunting of the crack tip and provides a significant decrease in the effective
stress intensity, a mechanical effect. The electrochemistry of the laminate
geometry also can interfere with the mechanism of stress-corrosion cracking
itself by altering the hydrogen concentration at the crack tip. It is be-
lieved that this is accomplished by the removal of the cathodic reaction,
either the decomposition of water or the reduction of hydrogen, and its con-
comitant molecular hydrogen or gaseous hydrogen from the crack tips to the en-
tire area of reacting interleaf, thereby decreasing the driving force for hy-
drogen cracking at the crack tip. Finally, two additional circwistances may
also lead to lessened corrosion at the crack tip. First, the cathode to anode
area ratio is small assuming that most of the exposed 30014 layer surface re-
mains active. Second, the 300M and especially the crack tip actually may t,-
come anodically passivated by reason of the galvanic corrosion. It should .-
noted that with the exception of some visible crack blunting at the stress-
corrosion crack tip the electrochemical arguments advanced above are specula-
tive, albeit plausible, and that their confirmation would require further ex-
perimentation and analysis of the electromechanical behavior of metal-metal
laminates.

84
 4.0 CONCLUSIONS

The present study has demonstrated and documented the feasibility of form-
ing complex component geometries from roll-bonded steel metal -etal laminate
plate stock by forging. In addition, it has been shown that the particular
laminate system studied composed of layers of 300M alloy steel separated by
interleaves of 1010 mild steel has distinctly improved fracture toughness,
fatigue strength, and, especially, stress-corrosion cracking resistance when
compared to monolithic 30014 steel.

Specifically, the following conclusions are justified:

* Roll bonding of steel alloys at high temperatures is an effective and

efficient process for fabricating thick multilayer lamilnates.

* Steel metal-metal laminates have limited but finite forgeability in the

longitudinal direction and excellent forgeability in the transverse di-
rection. The primary failure mechanism limiting formability in the
longitudinal direction is hot buckling.

9 In the forging of complex geometries from metal-metal laminates the die

design and lubrication are critical variables for successful forming.

* The tensile properties of laminates are controlled primarily by the

strength and volume fraction of the two component materials and secon-
darily by the strength of the inter-material interface and the relative
laminal thicknesses.

* Laminate fracture toughness and, in particular, the elastic-plastic

toughness or maximum resistance to critical crack growth is very much
superior to corresponding monolithic material in thick section.

85
* The improvement in laminate fracture toughness is primarily dependent
on the mechanical state of the laminate and the maintenance of a plane
stress or a mixed mode stress state during fast fracture. This is pre-
dominately a function of layer thickness and the ability of the layers
to decouple on fracturing, i.e., on the interleaf-layer bond strength
and interleaf plastic properties.
* There is an optimum value of bond strength for maximum fracture tough-

ness.

e The intrinsic fiber of metal-metal laminates provides a unique oppor-

tunity for controlling the directional properties of forgings using

laminates as forging stock.

* The fatigue strength of steel laminates has been shown to be superior

to monolithic material. This is primarily a consequence of the im-
proved damage tolerance of the laminates.

e The stress-corrosion cracking properties of 3001/1010 laminates have

been shown to be far superior to 3004 steel in simulated sea water. In
particular, the critical stress intensity for stress corrosion cracking
is more than double that of the monolithic steel.

86
 5.0 RECOMWNDATIONS

The results of the steel laminates properties characterization and forgea-

bility demonstration have provided sufficient experience and confidence that a
program for the fabrication of a prototype aerospace component based on steel
laminates now can be initiated. Such a prototype component program would re-
quire the sinking of a die insert specifically designed for the special attri-
butes of laminate forging, the fabrication of the component from laminate
parent material, and the testing and evaluation of the component. Properly

exploited the many exceptional properties of laminates, such as fracture

toughness, damage tolerance, fatigue strength, and stress-corrosion cracking
resistance would lead to weight savings and performance gains in aerospace
items of complex shape.
It would be desirable for the fatigue and stress-corrosion cracking pro-
perties of metal-metal laminates to be characterized and analyzed more fully
and in greater depth. In particular, the important parameters responsible for
the exceptional stress-corrosion cracking resistance of laminates need to be
defined more completely in specific service environments. This, then, would
allow the design of optimum laminate systems for specific applications and the
j extension of stress-corrosion resistance to other alloys and environments.
Such a study would provide, in addition, a novel approach for the fundamental
study of the mechanisms responsible for stress corrosion cracking.

87
 REFERENCES
1. Leslie Aitchison, A History of etals, Interscience, New York, 1960, in
two volumes, Vol. I, pp. 1-110.

2. J. R. Ellis and D. N. Hemanson, "Adhesively Bonded Multi-Layer F-104 Aft

Fuselage Ring Fitting," Vought Aeronautics Co., LTV Aerospace Corp., In-
terim Tech. Rept. IR-843-2(1), (1972).

3. J. R. Ellis and G. E. Kuhn, "Adhesively Bonded Multi-Layer F-104 Aft Fuse-

lage Ring Fitting," Vought Aeronautics Co., LTV Aerospace Corp., Interim
Tech. Repts. IR-843-2 (2-7), (1972-1974).

4. R. D. Goolsby, "Fracture and Fatigue of Diffusion, Explosive, and Roll

Bonded Al/Al and TI/Ti Laminates," ATC Report No. B-94400/7CR-23, AU-A040
387, prepared for Naval Air Systems Convaand on Contract No.
N00019-76-C-0288, (May, 1977).

5. S. V. Arnold, "A Laminated Specimen for Charpy Impact Testing of Sheet

Metal" Proc. Am. Soc. Testing Hats., 57, 1273 (1957).

6. S. V. Arnold, "Notch Sensitivity and Resistance to Tearing of Titanium

Alloy Sheet," Watertown Arsenal Laboratories, Tech. Rept. No. WAL TR
405.2/3, (1959).

7. S. V. Arnold, "Notch Sensitivity and Laminated Charpy Impact Strength of

1100-F and 2024-T6 Aluminum Alloy Simulated Sheet," Watertown Arsenal Lab-
oratories, Tech. Rept. No. WAL TR 341.5/1, (1959).

8. S. V. Arnold, "Toughness of Steel Sheet: The Advantage of Laminating,"

Watertown Arsenal Laboratories, Tech. Rept. No. WAL TR 843.21/2, (1960).

9. J. L. Bluhm, "A Model for the Effect of Thickness on Fracture Toughness

Applied to Laminated Structure," Proc. Am. Soc. Testing Mats. 61, 1332
(1961).

88

IIm
... . ._.. .._ _.._.. . ._
 10. H. L. Leichter, "Impact Fracture Toughness and Other Properties of Brazed
Metallic Laminates," J. Spacecraft, 7(3), 1113 (1966).

11. J. D. Embury, N. J. Petch, A. E. Wraith, and E. S. Wright, "The Fracture

of Mild Steel Laminates," Trans. AIME, 239 , 114 (1967).

12. J. G. Kaufman, "Fracture Toughness of 7075-T6 and -T651 Sheet, Plate, and
Multilayered Adhesive-Bonded Panels," J. Basic Eng., Trans. ASME, 89,
Series D (3), 503 (1%7).

13. R. F. McCartney, R. C. Richard, and P.S. Trozzo, "Fracture Behavior of

Ultrahigh-Strength-Steel Laminar Composites," Trans. ASH, 6U, 394 (1967).

14. E. A. Almond, N. J. Petch, A. E. Wraith, and E.S. Wright, "The Fracture of

Pressurized Laminated Cylinders", J. Iron Steel Inst., 207, 1319 (1969).

15. E. A. Almond, J. U. Emnbury, and E.S. Wright, "Fracture in Laminated Water-

5als," Interfaces in Composites, ASTM S P 452, Anerican Society for Test-
ing and Materials, Philadelphia, Pennsylvania, 1969, p. 107.

16. S. D. Antolovich, E. Kasi, and G. R. Chanani, "Fracture Toughness of Uu-

plex Structures: Part II - Laminates in the Divider Orientation," Frac-
ture Togness, Proceedings of the 1979 National Symposium on Fracture
Mechanics, Part II, AST STP 514, American Society for Testing and ater-
lals, Philadelphia, Pennsylvania, 1972, P. 1354.

17. S. D. Antolovich, G. R. Chanani, A. Saxena, and I. C. Wang, "Fracture

Mechanism Transitions in Laminate Composites," J. PIs. U: Appl. Ptws., 6,
560 (1973).

18. E. B. Kula, A. A. Anctil, and H. H. Johnson, "Fatigue Crack Growth in

Dual-Hardness Steel Armor," Army Materials and Mechanics Research Center
Technical Rept. AM4RC TR 74-6, (1974).

19. N. G. Ohlson, "Fracture Toughness of Laminated Steel,", Eng. Fract. Mech.,

6 (3), 459 (1974).

89

-L --
20. T. M. Devine, S. F. Floreen, and H. W. Hayden, "Fracture Mechanisms in
Maraging Steel-Iron Laminates," Eng. Fract. Mech., 6 (2), 315 (1974).

21. D. Cox and A. S. Tetelman, Improved Fracture Toughness of Ti-6AI-4V

Through Controlled Diffusion Bonding, Air Force Materials Laboratory Tech-
nical Rept. AFML-TR-71-264, (1972).

22. D. 0. Cox and A. S. Tetelman, Fracture Toughness and Fatigue Properties of

Titanium Laminate Composites Produced by Controlled Uiffusion Bonding, Air
Force Materials Laboratory Technical Rept. AFML-TH-73-288 (1973).

23. S. Floreen, N. Kenyon, and H. W. Hayden, "The Fabricability and Toughness

of Laminar Composites of Iaraging Steel", Trans. ASME, J. Engr. 14ater.
Tech., Vol. 96, Series H (3), 176 (1974).

24. J. F. Throop and J. J. Miller, "Fatigue Behavior of Metal Laminates", AD

785 690, Arm Science Conference Proceedings, Vol. III, AU 785 672, 229
(1974).

25. J. A. Alic, "Stable Crack Growth in Adhesively Bonded Aluminun Alloy Lami-
nates," Internal J. Fracture ,11 (4), 701 (1975).

26. H. W. Hayden and S. Floreen, "The Deformation and Fracture of Stainless

Steel Having Microduplex Structures", Trans. ASH, 61, 474 (1968).

27. S. Floreen, H. W. Hayden, and R. M. Ptlltar, "Fracture Behavior of an

Fe-Cu Microduplex Alloy and Fe-Cu Composites," Trans. AII4E, 245, 2529
(1969).

28. S. Floreen, H. W. Hzyden, and J. Q. Steigelman, "Fracture Modes in Lami-

nated Steel Nickel Composites," Trans. ASM, 62, 812 (1969).

29. J. Cook and J. E. Gordon, "A Mechanism for the Control of Crack Propaga-
tion in All-Brittle Systems," Proc. Roy. Soc. (London), 282A, 508-52U
(1964).

90

n n IN _noto
II I n &WAN_ -
 30. J. D. Embery, N. J. Petch, A. E. Wraith, and E. S. Wright, "The Fracture
of Mild Steel Laminates," Met. Trans, 239, 114-118 (1967).

31. J. F. Throop and J. J. Miller, "Fatigue Behavior of Metal Laminates,"

WVT-TR-75035, Watervliet Arsenal, June 1975.

32. J. F. Throop and R. R. Fujczak, "Fracture Resistant Titanium-Aluminum Lam-

inate," in Toughness and Fracture Behavior of Titanium, ASTTM STP 651,
American Society for Testing and Materials, Philadelphia, Pennsylvania,
1978, pp. 246-266.

33. R. M. Johnson, "Fracure and Fatigue of Diffusion, Adhesive, and Roll-

Bonded Aluminum, Titanium, and Ultrahigh Carbon Steel Laminates," ATU Re-
port No. B-921CO/BCR-80, AD-A058 553, Prepared for Naval Air Systems Gom-
mand on Contract No. NOU019-77-C-0287, Vought Corporation, Advanced Tech-
nology Center, May 1978.

34. R. M. Johnson and R. D. Goolsby, "Diffusion, Roll and Explosive Bonding of

Al/Al, Ti/Al, and Ti/Ti Laminates," in MATERIALS SYNERIAS 4i, 1UTH National
SAMPE Tech. Conf., Society for the Advancement of aterial and Prucess
Engineering, 1978, pp. 802-811.

35. L. E. Sloter and D. H. Petersen, "Fatigue and Fracture of Ultrahigh

Strength Steel and Titanium Roll-Bonded and Diffusion-Bonded Laminates,"
ATC Report No. R-92000/UCR-31, prepared for Naval Air Systems Command on
Contract No. N00019-78-C-0491, AD-A086 511/3, Vought Coporation, Advanced
Technology Center, May 1980.

36. L. E. Sloter and D. H. Petersen, "Fracture and Fatigue of Metal-etal Lam-

inates," in Material and Process Applications, Land, Sea, Air, Space, 26th
- National SANPE Symposium Proceedings, Vol. 26, Society for the Advancement
* of Material and Process Engineering, 1981, pp. 313-324.

37. "Standard Method of Test for Plane-Strain Fracture Toughness of Metallic

Materials," ASTM E 399-81, American Society for Testing and Materials,
Philadelphia, Pennsylvania, 1981.

91

4 m
AD-A127 541 ROLLBONDED 30DM/1010.STEEL METAL-METAL LAMINATES:
FORGEABILT TOUGNESS..U) VOUGHT CORP ADVANCED
TECHNOLOGY CENTERDALLAS TXS E SLOERJUN 82
UNCLASSIFIED ATC-R-92000/2CR 3D NODS 9-80-C 0575 16 N 8/
 IO

111 1.0 ~ lILI

28
--
W
LAiIiL-

MICROCOPY RESOLUTION
TESTCHART
NATIONAL
BUREAU
OF STANDARDS-
1963-A

4-.....
38. "Military Specification: Steel Bars, Reforging Stock, and Mechanical Tub-
ing Low Alloy, Premium Quality," MIL-S-8844C, 25 May 1971.

39. J. E. Jenson (ed.), Forging Industry Handbook, Forging Industry Associa-

tion, Cleveland, Ohio, pp. 193-194.

40. Alloy Digest, Engineering Alloys Digest, Inc., Upper Montclair, New Jersey.

41. Metals Handbook, Ninth Edition, Vol. 1, Properties and Selection: Irons
and Steels, American Society for Metals, Metals Park, Ohio, 1978, pp. 120,
125, and 127.

42. "Standard Methods of Tension Testing of Metallic Materials," ASTH E 8-81,

1981 Annual Book of ASTM Standards, Part 10, Metals - PtWsical, Mechani-
cal, Corrosion Testing, American Society for Testing and Materials, Phila-
delphia, Pennsylvania, 1981, pp. 197-217.

43. "Standard Test Method for Plane-Strain Fracture Toughness of Metallic Ma-
terials," ASTM E 399-81, 1981 Annual Book of ASTH Standards, Part 1U,
Metals-Physical, Mechanical, Corrosion Testing, American Society for Test-
ing and Materials, Philadelphia, Pennsylvania, 1981, pp. 588-618.

44. W. A. Backofen, Deformation Processing, Addison-Wesley, Reading, Massachu-

setts, 1972, pp. 151-152.

45. Atlas of Isothermal Transformation and Cooling Transformation Diagrams,

American Society for Metals, Metals Park, Ohio, 1977, pp. 396-397.

46. F. E. Adams, Vought Corporation, Dallas, Texas, private conmunication.

47. J. E. Jenson, Forging Industry Handbook, Forging Industry Association,

Cleveland, Ohio, p. 106.

48. F. A. D'Isa, Mechanics of Metals, Addision-Wesley, New York, pp. 100-102.

92
 4.F. R.Sane 81nelasti Colun Theory," . Aer.~ Sv., 14(5) 26-6

* (1947).

* I50. S. L. Semlatln and H. R. Plehier, "Deforiuatlon of Sandwich Sheet Materials

in Unlaxial Tension,6 Met. Trans., 10A, 85-96 (1979).

*51. S. L. Semlatin and H. R. Plehier, "Formability of Sandwich Sheet Materials

inPlane Strain Comprssion and Rollng, Met. Trans., 10A, 97-107 (1979).

5.F. A. D'Isa, Mechanics of Metals, Addislon-Wesley, New York, P. 143.

54. Irwin Miller and John E. Freund, Probability and Statistics for Engineers,
Prentice-Hall, Englewood Clifs, New Jersey, 1965, pp. 226-231.

55. J. E. Jenson, Forging Industry Handbook, Forvinq Industry Association,

Cleveland, Ohio. pp. 10-11.

56. Damage Tolerant Design Handbook, MCC-IO-01, Metals and Ceramics Informa-
tion Center, Battelle Columbus Laboratories, Columbus, Uhlo, January 1975,
pp. 7.2-3 (12/72) and 6.2-11(1/75).

.993

-ii
 I Page I

CONTRACT DISTRIBUTION LIST

Comwander
Naval Air Systems Command
Attn: J. S. Bruce (AIR-5163C6)
Washington, DC 20361

Commander
Naval Air Systems Command
Attn: Code AIR-950D
Department of the Navy
Washington, DC 20361

Commander
Department of the Navy
Naval Sea Systems Command
Attn: Code 03423 (1 copy)
Washington, DC 20361

Chief of Naval Research

Department of the Navy
Attn: Code ONR 423 (1 copy)
Code ONR 471 (1 copy)
Washington, DC 20361

IU. Director
S. Naval Research Laboratory
Attn: Dr. Ray Hettche
Dr. B. B. Rath
Dr. G. Yoder
Washington, DC 20390

Commander
Naval Air Development Center
Attn: Code 606
Waruinster, PA 18974

•. . Commanding Officer
Naval Air Rework Facility
Naval Air Station, Bldg 604
Attn: Code 34100
Pensacola, FL 32508

Comandilng Officer
Naval Air Rework Facility
Naval Air Station, North Island
Attn: Code 34100
San Diego, CA 92135

Naval Material-Industrial Resources

Philadelphia, PA 19112
* Page 2

Contract Distribution List (Continued)

Director
Air Force Materials Laboratory
• Attn: Codes LLMO (1 copy)
LC (1 copy)
MXE (1 copy)
NBC (1 copy)
LLS (1 copy)
Wright-Patterson AFB, OH 45433

Army Aviation Systems Command

P. 0. Box 209
St. Louis, MO 63166

Department of the Interior

Bureau of Mines
Washington, DC 20240

U. S. Energy Research & Development Administration

Division of Reactor R&D
Attn: Mr. J. M. Simmons (1 copy)
Chief, Metallurgy Section
Washington, DC 20545

Battelle Memorial Institute

Defense Metals Information Center
"- 505 King Avenue
Attn: Yr. Richard Wood (1 copy)
Columbus, OH 43201

AVCO Space Systems Division

Lowell Industrial Park
Lowell, MA 01851

Brush Wellman, Inc.

17876 St. Clair Avenue
Cleveland, OH 44110

NASA/Langley
Manufacturing Technology Section
Attn: Hr. Tom Bales (1 copy)
"1- Rampton, VA 23365
HNeDonnell-Douglas Research Laboratories
FP. 0. Box 516
! Attn: Dr. D. P. Ames (1 copy)
Dr. C. Whitsett (1 copy)
Mr. H. C. Turner (1 copy)
Mr. H. J. Siegel (1 copy)
(M&P Development Dept,
General Engineering Div.)
St. Louis, HO 63166
The Franklin Institute Research Laboratories
Twentieth & Parkway
Attn: Technical Director (I copy)
Philadelphia, PA 19103

* ~ * *>A a4.a
 Page 3

Contract Distribution List (Continued).

Dr. John A. Schey (1 copy)

Department of Mechanical Engineering
University of Waterloo
Waterloo, Ontario, Canada N2L 3 Gi

AiResearch Company
Materials Applications Group
93-3G1-503-4V (1 copy)
402 South 36th Street
Phoenix, AZ 85010

Wyman Gordon Company

Attn: Mr. Charles Gure (1 copy)
Worcester Street
North Crafton, MA 05163

Kawecki Berylco Industries

P. 0. Box 1462
Attn: Dr. J. P. Denny (1 copy)
Reading, PA 19603

Ladish Company
Packard Avenue
Attn: Mr. Robert Daykin (1 copy)
Cudahy, WI 53110

Linde Company
Division of Union Carbide
P.O. Box 44
Tonawanda, NY 14152

Midwest Research Institute

425 Volker Boulevard
Kansas City, MO 64110

Nuclear Metals, Incorporated

Attn: Dr. Paul Lowenstein (1 copy)
2229 Main Street
Concord, HA 01742

General Electric
Missile & SpAce Division
Materials Science Section
Attn: Technical Library (1 copy)
P. 0. Box 8555
Philadelphia, PA 91901

Reynolds Metals Company

Reynolds Metals Building
Attn: Technical Library (1 copy)
RIclmond, VA 23218

IL... ....
 Page 4

Contract Distribution List (Continued)

Artech Corporation
2816 Fallfax Drive
Attn: Mr. Henry Rahn (1 copy)
Falls Church, VA 22042

LTV Aerospace Corporation

SPVR, Engineering Materials
Vought Systems Division
Attn: Mr. A. E. Hohman, Jr. (I copy)
P. 0. Box 5907
Dallas, TX 75222

Advanced Technology Center, Inc.

Attn: Dr. D. H. Peterson (1 copy)
(Senior Scientist)
P. 0. Box 6144
Dallas, TX 75222

Defense Advanced Research Project Agency

1400 Wilson Boulevard
Attn: Dr. E. C. VanReuth (I copy)
Arlington, VA 22209

Grumman Aerospace Company

Advanced Material & Processing Div.
Attn: Mr. Carl Micillo (1 copy)
Bethpage, LI, NY 11714

I Aluminum Company of America

1200 Ring Building
Attn: Mr. G. B. Barthold (1 copy)
Washington, DC 20036

Massachusetts Institute of Technology

Dept. of Metallurgy & Material Science
Attn: Dr. N. J. Grant (1 copy)
Cambridge, MA 02139

Drexel University
Dept of Metallurgical Engineering
4 Attn: Dr. Alan Lawley (1 copy)
32nd & Chentnut Stri'tCt
Philadelphia, PA 19104

"I Reactive Metals. Incorporated

Attn; Dr. Howard Bomberger (1 copy)
Mle, OH 44446

Lockheed Aircraft
Dept. 74-50, Bldg. 985
Attn: Mr. Rod Siemens (1 copy)
urbank, CA 91520

' : * . •:, '. .

* ,. - .* . _ /
 Pate 5

Contr act Distribution List (Continuedl)

Soceed Cesste pc jIc

Attn: Dr. Neilmato (1 copy)I cpy

1049Ca 31o/Do04io
325ousanve OaStCA9136

Plo BAt, CA 9404

Divisiol ofIniteairaft Cororpto

Floidac Resterc&Delomn r
P. 0. Box 2691
4 Attn: Mr. JNillerto (1 copy)
Thesald Bac,CL 93302

Crugiles Aicatl Resarcey

P. 0. Box 889

Parkway West & Route 60

Attn: Mr. E. J. Dulis (1 copy)
Pittsburgh, PA 15230

Reynolds Metals Corporation

Attn: Mr. George Hsu (I copy)
Manager of Industry Standards
6601 West Broad Street
Richmond, VA 23261

Henry Xrumb School of Mines

Columbia University
Attn: Dr. John K. Tien
New York, NY 10027

Couild, lite.
Gould Information Center
540 East 105th Street
Cleveland, 0CH 44108
Comonealth Scientific
500 Pendleton Street
Attn: Mr. A. P. Divecha (I copy)
Alexandria, VA 22314
 6
I? Wage

Contract Distribution List (Continued)

The Boeing Company

12842 - 72nd Avenue, N.E.
Attn: Mr. W. Spurr (1 copy)
Kirkland, WA 98033

Rockwell International
Columbus Division
Attn: Mr. P. Maynard (1 copy)
Dept. 75, Group 521
Columbus, OH 43216

Rockwell International
Los Angeles Division
International Airport
Attn: Gary Keller (I copy)
(Materials Applications)
Los Angeles, CA 90009

Martin-Marietta Aerospace
New Technology - Metals
Denver Division
Attn: Mr. Geisendorfer (1 copy)
P. O. Box 179
Denver, CO 80201

Grumman Aerospace Corp

Adv. Mat. & Processes

J ATTN: Mr. Bill Grant

Bethpage, LI, NY 11714

___

* %iV
17
 * ---. <-..-----~----~-~

I
* ~ fl~ .~.&fr-,