i --- studies of the strain distribution in wide plates
TRANSCRIPT
n,.
-Jr, *,
SSC-118
--
I
---
I
II-
1,
I1’....,
,—
1’-.
.-
!J
STUDIES OF THE STRAIN DISTRIBUTION IN WIDE PLATES
DURING BRITTLE FRACTURE PROPAGATION
by
S, T, Rolfe
T. M. Lynam
and
W. J. Hall
, -3
I
!.—’
SHIP STRUCTURE COMMITTEE
II
COMMITTEE
MLMBER AGEUCiBB:
DUREAU OF SHIM. D-. or NAVY
MILITAW EM TRAMOWNTATION ●ERVICC,DEW. OF MAW
UNITED STATEO COA8T ~UARD, TmEAwRY DEmT,
MAWTIMK ADMINISTRATION. D@T. or COMMKRCS
AMCRICAN ■ungw or 8MIP?IMm
ADDRtBs CORRiHPW~
I 8RCRHAW
I SOIIP ,rwlorunc COMMIT?*K1,
)
. IIU. 9. COA97 9uAm0 HmAOWAWKR@ ii
I
I
December 30, 1959 ~
Dear Sir:I
As part of its research program related, to the improvementof hull structures of ships, the Ship Structure C~mmittee is sponsor-ing an investigation of Brittl@ Fracture M=han~~s at the univ@rs@of Illinois. Herewith is a copy of the Fourth Pr$gress Report, SSC-118, Studies of the Strain Distribution in Wide F!lates Durinq Brittle—— ——Fracture Propagation~S. T. Rolfe, T. M. Lyn~m, and W. J. Hall.
This project is being conducted under t~e advisory guidanceof the Committee on Ship Structural De siqn of the National Academyof Sciences-National Research Council.
This report is being distributed to individuals and groups!3associated with or interested in the work of the hip Structure Com-
mittee. Comments concerning this report are sblicited.
. Sincerely youts,
E. H. Thiele IRear Admiral, U. S. Coast GuardChairman, Ship Structure Committee
r-
!]l.,
-,
r-
—..
ni=;
,—I
,!
I—,
1~.—
‘1
I
Serial No. SSC-118
Fourth Progress Reportof
Project SR-137
to the
SHIP STRUCTURE COMMITTEE
on
STUDIES OF THE STRAIN DISTRIBUTION IN WIDE PLATESDURING BRITTLE FRACTURE PROPAGATION
—
by
S. T. Rolfe, T. M. Lynam and W. J. Hall
University of IllinoisUrbana, Illinois
under
Department of the NavyBureau of Ships Contract NObs-657’90
BuShips Index No. NS-731-O 34
transmitted through
Committee on Ship Structural DesignDivision of Engineering and Industrial Research
National Academy of Sciences-National Research Council
under
Department of the NavyBureau of Ships Contract NObs- 72046
BuShips Index No. INS- 731-O36
Washington, D. C.National Academy of Sc iences-Na.tional Research Council
December 30, 1959
ABSTRACT
This report summarizes the results of a series of tests made as
a part of the study of the propagation of brittle fractures in 6-ft wide
steel plates. All plates were tested at an average net applied stress
of 19, 000 psi, a temperature of about -10 1?, and an impact energy of
about 1000 ft-lb, which made it possible to superimpose the test data
and obtain contours of strain on the surface of the plate for a propagat-
ing fracture. Contours of both the maximum principal strain and strain
measured. with vertically oriented, gages for various lengths of crack are
presented in this report. A study of all the applicable data from earlier
tests made as a part of this program indicates that the strain contour
data presented here are also representative of the data from these earlier
tests. The studies indicate that for the particular specimen geometry
and associated test conditions, the strain field associated with the tip
of the advancing fracture remains essentially unchanged after traversing
about one-third of the plate width, and extends only about 8--10 in.
ahead of the crack tip.
—
CONTENTS
—.
INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . .
General . . . . . . . . . . . . . . . . . . . . . . . . . . .
Object and Scope . . . . . . . . . . . . . . . . . . . . . .
Acknowledgment . . . . . . . . . . . . . . . . . . . . . .
Nomenclature . . . . . . . . . . . . . . . . . . . . . . . .
DESCRIPTION OF SPECIMENS AND INSTRUMENTATION. .
Specimens and Material Properties. . . . . . . . . .
Instrumentation . . . . . . . . . . . . . . . . . . . . . . .
Data Reduction . . . . . . . . . . . . . . . . . . . . . . .
Apparatus and Test Procedure. . . . . . . . . . . . . .
ANALYSIS OF TEST RESULTS . . . . . . . . . . . . . . . . . . .
General . . . . . . . . . . . . . . . . . . . . . . . . . . .
Recorded Test Data . . . . . . . . . . . . . . . . . . . .
Computed .Principal Strains . . . . . . . . . . . . . . .
Discussion of Strain Traces . . . . . . . . . . . . . . .
Maximum Principal and Vertical Strain Contours . .
Crack Path and Surface Texture. . . . . . . . . . . . .
SUM~Y . . .. o . . . . . . . . . . . . . . . . . . . . . . . . .
REFERENCES . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
1
1
1
2
3
3
3
6
17
17
18
18
20
39
41
45
60
62
64
BRITTLE FRACTURE MECHANICS ADVISORY COMMITTEE
for the
COMMITTEE ON SHIP STRUCTURAL DESIGNDivision of Engineering & Industrial Research
National Academy of Sciences-National Research Council
Chairman:
N. J. HoffHead of DepartmentStanford University
Members:
D. S. Clark
of Aeronautical-Engineering.-.
—
Professor of Mechanical EngineeringCalifornia Institute of Technology
Morris CohenDepartment of MetallurgyMas sachusetts Institute of Technology
F. J. Feely, Jr.Esso Research & Engineering Company
Martin GolandVice PresidentSouthwest Research Institute
G. R. IrwinHead, Mechanics. DivisionNaval Research Laboratory
Egon OrowanDepartment of Mechanical EngineeringMassachusetts Institute of TechnoJ.ogy
W. R. OsgoodDepartment of MechanicsRensselaer Polytechnic Institute
M. P. WhiteHead, Civil Engineering DepartmentUniversity of Ivlassac”husetts
—
—
INTRODUCTION
General
Brittle fractures in riveted and welded steel structures have been re-
ported in the engineering literature for many years. These fractures generally
are characterized 1) by a lack of the ductility usually associated with failures
of structural steel and 2) by a sudden occurrence with little or no previous warn-
ing.
The significance of the brittle fracture problem was not fully appreciated
until World. War II, when a large number of welded merchant vessels failed in
this manner. Fortunately, through the use of improved geometrical layout, crack
arrestors, and improved materials and fabrication procedures, it was possible to
reduce greatly the number of major ship failures. In many cases, provisions have
been made to incorporate similar improvements or changes in the design of struc-
tures other than ships in order to minimize the possibility of the occurrence of
brittle fractures. Neverthele SS, in spite of these improvements in design which
have re suited from the large amount of research completed during and since World
War 11, brittle fractures still occur, and further studies are requirtid if a better un-
derstanding of the
-.-t a@ QE!Q
The brittle
a consideration of
brittle fracture problem is to be obtained.
fracture phenomenon is extremely complicated in that it involves
materials and their behavior in various environments. This par-
ticular program is concerned with a study of the propagation of brittle fractures in—
wide steel plates. The experimental approach followed in this investigation has
consisted of measuring surface strains and crack speed as the fracture traverses a
wide steel plate to obtain fundamental data.
The primary purpose of these tests and studies was to obtain sufficient
data to establish representative strain contours on the plate surface during the time
a brittle fracture is propagating. Thus far, six major fracture tests of 6-ft wide,
semiskilled steel plates have been completed in which 34 channels of cathode ray
oscilloscope recording instrumentation have been utilized; in each test, 33 channels
were used for 11 rectangular strain rosettes and one channel was used for speed
-2-
detectors. The plates were tested at an average net applied stress of 19, 000 psi
and at a temperature of about -10 F. The fractures. were initiated, from one edge
of the plate with the notch-wedge-impact method of fracture initiation. In con-
junction with the crack arrestor program (Project SR- 134), one additional set of
records has been obtained for a plate tested at a higher stress level (28, 000 psi)
and at a lower temperature (-15 F). The results presented in this report are based
on the tests that are reported here for the first time, as well as on all of the pre-1-4
vious applicable data obtained as part of this program.
Among the more important items presented in this report are the following:
—
—
(1)
(2)
(3)
(4)
{5)
The
re suits and
cal results,
strain gage traces from the component gages of the strain rosettes
principal strain curves computed. from the component gage strain traces
discussion of the factors affecting the computed principal strain magni- –
tudes and strain rates
representative vertical and maximum print ip al strain contours
ous crack lengths during the time the fracture is propagating,
typical sets of maximum print ipal strain contours and vertical
for vari- A
and
strain
contours associated with a crack in the central portion of t“he plate.
strain contours should be of considerable value in correlating the test
associated significant parameters with other experimental and analyti-.-
Acknowledgment
The work described. in this report was conducted. in the Structural Research
Laboratory of the Department of Civil Engineering, University of Illinois. The
project is under the general direction of N. M. Newmark, Professor and Head of
the Department of. Civil. Engineering. T“he program is. sponsored by the Ship Struc–
ture Committee, and the members of the Brittle Fracture Mechanics Advisory Com-
mittee under the Committee on Ship Structural Design of the National. Academy of
Sciences-National Research Council have acted in an advisory capacity in the
planning of the program.
v. J. McDonald, Associate Professor of Civil Engineering, supervised the
instrumentation and reduction of the test data from the photographic records. M.
—
P. Gaus, Research Associate in Civil Engineering, prepared the computer code for
-3-
—
.
.
t“he strain rosette computations, and F., W. Barton, Research. Assistant in Civil
Engineering, helped with the tests and preparation of the figures for the report.
Nomenclature
The following terms are used repeatedly throughout the text:
Dynamic strain gage-- SR-4 Type A7 ( l/4-in. gage length) strain gage whose
signal is monitored, with respect to time on an oscilloscope during the
fracture test.
Static strain gage--SR-4 Type A7 ( l/4-in. gage length) strain gage used to
monitor the static strain level.
Component strain gage-- One of the three. individual strain gages of a rectan-
gular strain rosette.
Crack detector--A single-wire SR- 4 Type A9 (6-in. gage length) strain gage
located on the plate surface perpendicular to the expected fracture
path. A rough measure of the fracture speed may be obtained from a
knowledge of the distance between the detectors and the time interval
corre spending to the breaking of adj scent detectors.
Initiation ed,ge-- The edge of the specimen at which the brittle fracture is ini–
tiated.
Notch line--An imaginary straight line connecting the fracture initiation notches
on opposite edges of the plate specimen.
Test load strain--At any gage point, the strain corresponding to the applied test
load.
DESCRIPTION OF SPECIMENS AND INSTRUMENTATION
Specimens and Material Properties
The plate material for the six major fracture tests outlined in Table 1
(Tests 33, 34 and 36 through 39) was a semiskilled steel, USS heat No. 64 M487,
and was tested in the as-rolled condition. The steel was from the same stock
that was used in the earlier tests made as a part of this program and. its proper-
ties, typica~ of A– 7 structural steels, are in agreement with those reported ear-
lier.1,2,5
The specimens were 3/4 in. thick, 72 in. wide, and 20 in,. long;
“ 4-
TABLE 1. OUTLINE OF TESTS
.—
Test No. Initial Avg. Stress on Avg.(Plate No. ) Load Net Section TernP. Remarks
(kips) (ksi) (F)
With the exception of Test 35, all tests were conducted on 3\4-in. by 72-in. by 120-in. semi-skilled steel plate specimens welded with E7016 electrodes to l-in. thick pull-plates in the 3, 000, 000-lb Baldwin hydraulic testing machine. The brittle fractures were initiated by the notch-wedge-impactmethod with a nominal lateral impact of 1.200 ft-lb,
The notch was 1- 1/8 in. long and consisted of a slot four hacksaw blades wide (-O. 141 in. )for the first I in. , one blade wide(-o. 034 in-) for the ne~ lil 6 in. ~ andendedWitha jeweler’s saw-Cut (-0.012
;3C-1)
34(xF-l)
~;TRT-4)
(Tested inconjunctionwith ProjectSR - 134)
;;2B)
37(X2F)
38(XIB)
39(x2E)
., .,..in. ) l/lb in. long.
1000 19.0 0
1000 19.0 0
1475 28.0 -15
997 19,0
997 19.0
997 19.0
997 19.0
-lo
-8
-9
-6
Complete fracture--good strain records obtained from7 rosettes.
Complete fracture--g~od strain records obtained from10 rosettes.
Plate specimen composed of 36-in. starter strake ofrimmed steel, 4-in. strake of T-1, 20-in. strake ofrimmed steel, and a 12-in. strake of T-1 steel.Specimen was 27 in. long. Fracture arrested at lead-ing edge of final T-1 strake. Final load--85 kips.Good strain records obtained from 4 rosettes.
complete fracture--fair strain records obtained from11 rosettes. Double fracture last two-thirds ofplate width.
Complete fracture--good strain records obtained from11 rosettes.
Complete fracture--good strain records obtained from11 rosettes.
ComPlete fracture--good strain records obtained from11 rosettes.
—
-.
-5-
—.
Specimen Ready for Testing—.
—
—.
(!PULL
HEAD
I“ PuLLPLATE
SPECIMEN
l“ PULLPLATE
LPULLHEAD
?
t7’6”
-4
3’ a“
i
32’8”10’ o“
--l4’0”
i, ,,
J_
Lw+l
Line Diagram ofSpecimen
Fig. 1. Typical Test Setup
the net width at the notch line was approximately 2-1/4 in. less than the gross
width because of the notches on each edge. Both ends of the specimen were
welded with double-V butt welds made with E7016 electrodes to l-in. thick
pull-plates mounted in the testing machine; in welding the specimen to the
pull-plates, care was taken to keep the warping and residual stresses to a
minimum. A line diagram of the specimen and pull-plates and a view of
cal test setup in the 3, 000, 000-lb hydraulic testing machine are shown
Fig. 1.
The composite plate specimen for the arrestor test (Test 35) was
a typi-
in
fabri-
cated from, in order, 1) a 36-in. fracture starter strake of rimmed steel (the
plate on which the strain rosette gages were mounted), 2) a 4-in. strake of
T-1 steel, 3) a 20-in. strake of rimmed steel, and 4) a 12–in. strake of T-1
steel. Over all, the specimen was 3/4 in. thick, 72 in. wide, and 27 in.long.
Mechanical property tests were made on material taken from the cen-
tral portion of the plates after the plates had been fractured. The check ana -
-.
n—
=
I
120I
TEST33 (XC–I)5EM1-KILLED STEEL
no
./
40
0
120I
TEST 84 (XF–11SEMI-KILLED STEEL
80
40
0
00I
TEST 39 (RTRT ’41RIMMED STEEL
40 /
o
ii 120m I
4 TEST 36 (X281SEMI-KILLEO STEEL
00/
. .: .
40
.,
:40 0 40 00 120 160
TEMPERATURE — deg. F
I
TEST 37 [XZF1SEMI-KILLED STEEL
80
40
0
1201
TEST 38 Cx I B)SEMI-KILLED HEEL
00 -
40
0
120I
TEST 3S 1X2,?)$EUI-KILLEO STEEL
ao
. . .
40
/
n
—,
.
—
~40 o 40 00 120 160
—TEMPERATURE — deg. F
Fig. 2. Charpy V-notch Curves --Tests 33--39
lys.es and the tensile test data for each of the rimmed and semiskilled steel plates
are presented in Tables 2 and 3. Charpy V-notch data are presented in Fig. 2.
Instrumental ion
Until recently, it had been possible to obtain only a limited amount of
data from any one test because a maximum of only nine channels of cathode-ray
oscilloscope recording ::quipment had been available. Nevertheless, earlier
work on this program with both 2-ft wide and 6-ft wide plates produced much
valuable information concerning expected fracture speed and the strain distribu-
. .
-7-
.-
TABLE 2. CHECK ANALYSES OF STEEL PIATE MATERIAL
Test No. Material Chemical Composition in Per Cent
c Mn P s Si Cu Ni Al
33 Semiskilled 0.17 0,71 0.019 0.028 0.058 0.02 0.0 0,03
34, 37 Semiskilled 0.20 0.76 0.019 0.028 0.052 0.02 Trace 0.03
35 Rimmed 0.18 0.4’2 0.013 0.031 0.02 0.23 0.14 0.003
36, 38 Semiskilled 0.21 0.82 0.018 0.030 0.05s 0.02 0.0 0.03
39 SemikiUed 0.20 0.76 0.019 0.040 0.03 0.04 0,16 0.002
.
--
TABLE 3. TENSILE TEST DATA FOR STEEL PLATE MATERLKL(Standard ASTM O. 505-in. diam specimens)
Test Material Heat Lower Yield Maximum ElongationNo.
Reduction(Plate No. ) No. Strength Strength in 2-in. of Area
(ksi) (ksi) 7’0 ‘%
33 Semiskilled(xc-1)
64M487 (L)* 32.9(T)**32.6
59.358.8
40,541.0
66.559.5
34 Semikillecl(xF-l)
64M487 (L) 32.9(T) 34.5
61.862.4
40.740.0
66.361.7
—.
.--
.—
35 Rimmed(ZIA)
16445 (L) 34.7(T) 35.2
68.168.7
36.531.2
57.651.6
36 Semiskilled(X2B)
64M487 (L) 36.5(T) 35.2
67.267.8
36.034.5
63.558.8
64M487 (L) 35.5(T) 35.0
67.861.8
37 Semiskilled(X2F)
64.464.2
41.036.3
38 Sernikilled(XIB)
64M487 (L) 35.5(T) 35.6
66.866.4
36.536.8
64.359.8
64M487 (L) 34.3(T) 35.2
61.662.1
65.357.5
39 Sernikilled(XZE)
39.536.8
*(L) Average of two specimens taken parallel to the direction of rolling.**(T) Averageof two specimenstaken transverse to the direction of rolling.
—
...
-8-
[
5PEEDOMAX
r
CALIBRATING OSCILLATORS
FFRANKLIN OS CILL05COPE5
r
OUMONT OSCILLVSCQPE
Fig. 3, Recording Equipment
tion at various points on the plate
surface as recorded by single
strain gages. The results of the
work through the end of 1956 have
been reported in a paper5 and, in
more detail, in two technical re-1, 2
ports . In 1957, an additional
25 channels of cathodeway oscillo-
scope recording equipment were
made available to the pr~gram by
the Naval Research Laboratory.
The latest 25 channels of
the cathode-ray oscilloscope
equipment were five-channel units manufactured by Franklin Electronics. The
nine earlier channels consisted of two DuMont Type 333 and two DuMont Type
322 dual-beam cathode-ray oscilloscopes, and a Tektronix Type 512 single
beam unit. The recording equipment, along with the calibrating oscillator that
supplied the time signals, is shown in Fig. 3; the car~eras for the DuMont
equipment are not shown.
Six of the DuMont oscilloscope channels are sufficiently sensitive to
allow at least 1-1/2 in. of trace deflection for O. 001 in. /in. of strain. How-
ever, the deflection scale used to record the test data on these DuMont scopes
was limited by the size of the scope face and the maximum value of expected
strain. The other 28 channels only allowed about 1/2 in. of trace deflection
for the same strain. In planning the recording of the test data, the equipment
was arranged so that the strain gages that were expected to produce signals ,of
lowest electrical magnitude were connected to the six DuMont oscilloscopes as
they had the highest sensitivity. The frequency response of all oscilloscopes
was flat up to at least 50 kc and therefore, they were considered to be adequate
.
—
to record the strain signals. Portions of two typical strain records are presented
in Fig. 4. The oscilloscope channels used to measure and record strain were
“9-
-.
—
Franklin oscilloscope Record
—..
—.
—
,.—
I?ig. 4. Typical Strain Records
calibrated by shunting gages with a resistance whose equivalent strain value
WEIS known.
The time base for the records obtained from the DuIWont and i%arddin
oscilloscopes was supplied by continuously moving 35-rrIm film.. Timing marks
orI the traces from the Franklin units were energized by the oscillator and re-
corded as part of the strain traces; the D-Mont units empl~yed intensity modu-
lation of the electron spot to define the time base. OrI all of the strain traces
a simultaneous “blip” was produced. by a synchronizing signal supplied from
the Tektronix oscilloscope circuit, immediately before and after the test. This
signal made it possible to synchronize accurately with each other all the strain
traces from a particular test. The synchronizing “blip” is visible orI the records
shown in Fig. 4.
The crack detector trace on the Tektronix oscilloscope was recorded on.
a single frame of W-mm film. The time axis was calibrated by putting a time
signal of known frequency on the channel and photographing one sweep. This
was done immediately after the test was completed.
Since only 34 channels of strain recording equipment were available
{ 3?J channels for 11 rectangular strain rosettes and one channel for crack speed
detectors) and the exact location of the fracture path was unknown prior to each
.—
-i O-
ri I. ROSETTE — GAGE LAYOLIT A [sEE FIG. 22)
t I I
DYNAMICSTRAIN GAGES CRACK PATH
Rosette Strain Orien- X Y Test J.OEd Distance tcNo. Gaw tation (in.) (in.) Strain Fracture
No. (in. /in.) (in.)
A 1. v !22.0 o .00067 0.-(; H 22.3 0 -.00019 0.8 113 D 22.5 0.5 .00024 0.5
B v 22.0 5.0 .00067 -2.35 H 22.3 3.0 -.00019 -2.2
D 22.3 3.3 .0$.024 -2.6c !V 22.0 20.0 .00069 -19.0
a E 22.3 20.0 -.00018 -19.09 D 22.3 23.3 .00026 -19.0
D 10 v 29.0 3.0 .00049 .1.111 H P7.3 3.0 -.00019 -1.012 D 2+.5 i3 .00026 -1.3
F 16 v 36.0 3.0 .00069 -0.517 H 36.3 3.0 -.00018 -0.518 n 36.3 3.5 .0002g -0.8
T 25 v 22.0 0 .00052 0.726 H 22.3 0 -.00017 0.8
(i:.) (ii.)
1.1 09 -0.2
1518 0:119 0.222 0.724 1.22s.5 1.936 2.545 3.248 3.454 3.5
3.72 3.9@ 3.970 3.872 z 6
27 D 22.3 0.3 .00016 0.5J 29 v 36.0 0 .00053 %.5
29 H 36.3 -. Ooole 2.6 IIw D 36.3 0!3 .00019 2.3
“IG 5 INSTRUMENTATION LAYOUT AND CRACK PATH – TEST 33
test, it was decided to concentrate the strain rosettes in three general areas
with respect to width orI three of the plate specimens, and thereafter to super–
impose the data to obtain the picture of the strain distribution associated with
a propagating crack.
The instrumentation layouts and crack paths for Tests 33 through 39 are
shown as Figs. 5-–11. ‘The strain rosettes are located on only the first two-
thirds of the plate, since it was felt that rosettes located in this region would
yield the desired strain information. The strain rosettes were located at 7-in.
—
.-
—
I
1~ 72” %
T[
.ROSETTE —GfJGE LkyOuT A (SEE FIG. 22)
I‘0.4m
‘o“I-1
.1
*D .H—’
.C
+-73&
=- B Fand K _ — — * ->
(DETECTORS)
—9’’LIO’’4L75LL5’L 7“4’
DYNAMIC STRtIN WOES CRACK PATd
Rosette Strain Orien- X Test Load Distance to:{0. Gage tation (in. ) (i:. ) Strain Fracture
No. (iii./ilL) (h. )
A 1 v 22.0 0 .00069 +3. 62 E 22.3 0 -.00010 +3. 63 D 22.3 +0.3 .00023 +3.3
B k v 29.0 0 .00066 +4.4
(it.) (ii.
1.3 04 .26 .4
12 1.5H ~,> -.00018
~ D 29.3 +0!3+4.4 16,5 2.5
.00326 +4.1 Mc
2.6v 29.0 +3.0 .00065 +1.4 22 3.6
; H 29,5 +3.0 -.00018 +1.4 24 3.8+~ 25.3 +3.3 .Goal? +1.1 =9 4.4D v 29.0 +6,0 .00069 -1,6 % !$,8
11 H 29,3 +6.0 -.00318 -~.6 42 5.512 D 29.3 +6.3 .0002& -1.9 48
E 13 vj.6
29.0 +9.0 .CO067 -4,6 54 5,814 E 29.3 +9.0 -.00018 -4,6 h i,o15 D 29.3 +9.3 ,00027 - !. 66 6,1
F 16 v 36,0 0 .00066 +4.8 72 6.117 E 36,3 0 -.00018 +4,818 D 56.3 +0.3 .00027- +k.5
H 22 v 36.0 +6.0 .00066 -1.223 ii 36,3 +6.0 -.00018 -1.224 D 36. T +6,3 @oo15 -1.5
I 25 v 36.0 +9.0 :00067 -4.226 H 3:.3 +9.0 -.00019 -4.227 D 3.5 +9.3 .00019 -4.5
J 23 v 22,0 0 .00057 +3. 6v 11 22.3 0 -,00017 +3.6
D 22.3 +0,3 .00021 +3.3K 31 v 36.0 0 .00057 +4.8
J2 H 36.3 0 -.00017 +4,833 D 36.3 ia.3 .00021 +1+.5
[G. 6 INSTRUMENTATION LAYOUT AND CRACK PATH – TEST 34
,
[- ‘M ‘“ ‘=, ROSETTE - OAOE LAYOUT A (sEE FIQ. 221
‘o
AOd Q a _c — >
,r~,iatl,.l.i
(DETECTORS)
‘*
RIMMED T-1 RIMMEO T-1
It
~ 36”I 4’4— 20”~ @4
DYNAUICSTRAIN GAGES
Rosette Strain Orieu- X Test Uaa Di~tame to
m s Gage tation (in. ) (il.) Str.2in FractureNo. (in./in.) (in. )
A 1 v 17.0 0 .(?0092 -0.55 E Ii, 3 0 - .c0024 -0.53~ .000 -0.8
B v 20.0 0 .00093 -0,6H a).3 -.00022 -0.6
~ D 22.3 +0!3 .0Q036 -0.9c v 23.0 0 .00092 -0,8
in 23.3 0 -.00022 -0.89 D 23.3 +0.3 .00039 -1.1
D 10 v 17.0 0 ,oC094 -0.5U H 17.3 0 -.00026 -0.512 D 17.3 +0,3 ,0002$ -0,8
CRACK PATS
(13,) (i:. )1.1 0
11 -.214.5 -.317 5a ~:623 -.8
=’1.1z -1.140.444 ::?48.5 053 +.357 060 -1.462 Fractore
arregtedin T-1
FIG. 7 INSTRLFAENTATIQN LAYOUT AND CRACK PATH - TEST 35
Y_72”~
[ ‘- < K ‘ ‘ “ ‘-2
,ROSETTE - GAGE LAYO!JT K iSEE FIG. 221
‘o ,7 .0
.J
.6 4 .5
2 3
— 9_ rO _— L— —— x
‘0
“mi--r$’+l+ 7“:7’’+7”’$7” ,~: ,,,,+’ ‘DETECT”’”
DYHMUC STBAIN GAGES
Rosette strain Orien- X Y Teet bad Di6t,anceto
Ho. Gage tation (In.) (In.) Strdn FractureMa. (in./ino) (in.).,.
1 1 ri 2240 3.0 -.00016 0.O (bwer)
2 D 22.3 3.3 .00022 0.0 “
27 D 2qm3 3.3 .GO026 o.% “a v 2io 3.3 .00074 0,5> 1’
3 11 El 36.0 3.0 -, Wol? 1.35 “12 D .00027 1.80 “1 . 00o-fo 1.70 “
h 1, -.00016 0.0 (Upper)IT D G:: :.3 . 00QZI 0.10 “18 29v .Ooo-11 0.10 “
5 a H 6:0 -.00017 2.6) “22 D ~;; 6.3 .00026 2.15 “25 6,3 .00069 2.2G “
6 4; 22:3 .00017 -3.00 (Lower)
5 v 22.0;:;
.00070 -3.10 “
24 H 22.0 6.0 -.00016 -2.80 “7 26 R a.o 9.0 -.00015 - ~r
ij D 2;.5 j.3 .03026 -3. J.O ‘128V w 9.3 .00372 -3. ?.0 “
8 ~ H 36:8 9.o -, 000IT -0.45 “32 D 3$: 9;; .oca2-/ -0.80 “31V. .00071 -0.75 “
9 9~ 22,5 0,3 .Oooa 5.05 (Im+2r)
10 v 22.0 0.3 .00069 2.90 “
25 H 22.0 0.0 -.00017 3,25 “10 14 D 29,3 0.3 .000% 2.55 “
15 v W.o 0.3 .00063 2.55 “—.30 H 2j.O 0.0 -.00017 2,85 “
11 19 D 36.3 0.3 .00027 1.15 “20 v 36.0 0.3 .ODOTO 1,15 “33 H 36.0 0.0 -.00017 1.65 “
CMCK PATS
x.Ih) (i:.}— —1.1 05 .2
LO ,9L5 1.9a 2.622 ?).2
25.0* 3.8(u) (L)
?6 4.2 %4ZT 4.6 3.1s 5,9 2.8m 6.5 2.735 8.1 1.933 8.7 1.2i5 10.3 0.743 11.3 0.352 11.0 0.355 12.1 0.357 12.5 059 12.9 “0.465 13.5 -0.966.5 13.5 -LO$,5 ;;.; -0.8
.1.072 13:6 -O. T
Crack Brtmche&U)= Upper Crackl,)=Lower Crack
‘1(38 INSTRUMENTATION LAYOUT AND CRACK PATH – TEST 36
?25’ ~
~y~OSETTE - GAGE LbYOIJT B [SEE FlG,22)
‘0
4 ~ ‘- ‘ - ‘
.7“m
-e .6 .5
.4 .2 .1
— 9_ .11— ——.3 —
do‘o
-m
75’’+ 7’’+7+7 +7+ ,~ ,,,,~F’DE’E’TORS’
DYNAMICSTRAIN GAGES
Iosette Strain Orien- X Y Test lad Distance toNo, Gage to.tion(in.) (in,) Strain Fr.2cture
Ho. (in./in,) (in.)
1 1 H 22 3 -.0C018 -1. To2 D 22.3 3.3 ,00022 -2.103 v 22 33 .00063 -2.10
2 6 H 15 3“ -,00018 -2.40D 15.3 3.3 .00022 -2.70
1 T 15 00066 -P,703 11 H 15 -3 -.00018 +1.65
12 D L5.3 -2.7 .00023 +1. 3513 v 15 -2.7 .00065 +1. 35
4 16 110 3 - .ocol~ -2.!30l-l D 8.3 3.3 .00023 -3.1018 Va .00067 -3.10
5 a E 36 6’ -.00017 -3.6522 D 3;.3 6,3 .00024 -4.0023 v 3 6,3 .00067 -4.00
6 24 H 22 6 -.0001’7 -4.704 D 22.3 6.3 .Ocwzl -5,105 v 22 6,3 .ooa67 -:&
7 26 H 22 -,00017 -i .-102-7 n 22.3 ;.3 .00027 -8..1023 v 22 .00066 -8.10
8 29 H 15 6“ -, CO018 -5.4032 D 15.3 6,3 .00026 -5.7031 v 15 6.3 .00069 57 0
9 25 H 22 0 -.oc016 “ ‘+1. 359D 22.3 .3 .00027 +1.00
10 22 3 .Ocofo +1,0010 3DH8 -.00017 +2. m
14 D 8,3 :;.7 ,0001$ +1. p15 V 8 -2.7 .oo36\ +1.9c
U 33 H 36 0 - .C#l-l +2. 351P D 56.3 .00023 +2.00Eo v 36 :; .00066 +2.00
CSACK PATH
(i) (i:.)1.1
2225
2932,535.536Y?
697072
00.2.6
1.11.61.82,22.62.62.42.43.53.6
.
;:;3.63,2
FIG, 9 INSTRUMENTATION LAYOUT AND CRACK PATH – TEST 37
I I 1
II
I 1i
Y72”~~. .
I
~
- 2 4
MOSETTE – GAGE LAYOUT B tSEE FIG 221
I
.7 .8
‘y .4 ,6
“O.5 . .1
.3 .10 .9
f = 1-
L——— — — —x
‘0b B co E F {DETEcTORS]
DYNAMIC STFWH GAGES
R09ette Strain Orien- X Y Test bad Distance toHo. Gage tation (in.) (h.) Strain Frectwe
No. [lni/ino) (in.).,.1 1 H50 -.00017 -O.-1o
2 D 50.3 :,3 .@oo13 -1.053 v jO 6.3 .00046 -1.05
2 H 45 - .9001a -1.95-r D 43.3 6.3 .00019 -1. i8 v 43 6.3 .00056 -low
3 11 H 56 3 - #50017 +]. .6512 D 56.3 3.3 ,00018 +1. 2515 v 36 3.3 .90050 +1, 25
k 16 il 43 9 -.00019 -4.0517 D 43.3 9.3 .00021 -4,33
— 18 v 43 -9.3 .00056 -4.4C5 ‘a n 35 -,00017 -1471
22 D 36.3 6.3 .coo14 -1.732> v J6 6, 3 .OGGEO
6a
24 z 50 9 -, GOG174
-3.70D 50.3 9,3 .00015 -i.05
5 v 9.3 .00056 -Lo 57 26 H 43 12 -,00015 -7.05
27 D 1+3.3 12.3 ,00013H
-7.3v 4> 12.5 .00044 -,740
a 29 H 50 12 -.00013 -6,-K)32 D 50.5 12.3 .00023 -7.0551 v >0 1.2.3 .0GQ57 -7.05
9 25 E5Q -, 0001-( +2, 359 D 5Q,3 ;.3 ,Oooa +2. 10
10 v 50 3,3 .00056 +2.0010 n H 43 -,00017 +1.95
lb D 43.3 ;. 3 ,~f.)ps +1. 7015 v 43 5.3 .00056 +1. 65
11 33 H o -.00018 +4,65E) D 36.3 ,3 .DDO09 +1), g20 v 36 .3 .000>4 +4. ~
CRACK PATH
(L1,) (i;.)
1.1
;::10
222427m3540b5505456m657072
FIG. 10 INSTRUMENTATION LAYOUTAND CRACK PATH – TEST 38——— ——.
0,2
,3
,82.93,25.53. i?’4,64.97.15.45.45.35.45.55.55.4
{ , “‘“r ~-?‘y .7 ,G
-m ,6 .4,5
.1 .2 .3
~__. 10 __ .1!— — — — —)
‘0
“m_-75”; 7L’;7’&7,,&7L, ,~: ,,,,~’ “’’’’’o’”
DYHAWC S~IN GAGES
Rosette Strain Orien- X Y Test Load Distance toNo. ihge tation (In. ) (in,) Stmln Fracture
Ho. (inB/in.) (in,)
1 1 H “22 3 - .moli’ +2. 292 D 22.3 3.3 .CO021 +2. 103 v 22 35 00062 +1. q6
2 6 H a 3’ - .0CQ17 +4.147 D 29.5 3!3 .00019 +3.938 v % 3.3 .cmo61 +3.85
3 11 H 36 -. CO017%.3 ;.3
+4.5012 D . ooQ19 +4. 2213 v 2 coo%
4+4.12
~6 li29 6“ -.00016 +1 .0917 D 29.3 :.3 ,00022 +0.8918 Y .CO063 +0.85
5 ‘a. K % 6“ -. 0001-/ +1. 5522 D %.> 6.5 .00023 +1, i4
3 36 6,32 v6
.cQo62 +1, 1224 H 22 -.00017 -0,73~ D 22.3 2,3 .oco25 -0. go5 v 22 0005 6 -1,08
7 26 H 29 -.00018 -1.7627 D 29.3 ;,3 .Cooz-o -2.07
v .000618
-2,12N H 36 9“ - [email protected] -1.4?32 D %43 9m3 .CW223 -1.8031 56 9.~v .co062 -1..8
9 25 E22 o - .00J17 +5.279D 22,3 0.3 .0C022 +5. 10
10 v 22 0.3 .m59 +4.9810 m E 29 - .m19 +7.17
14 D 29.3 0:3 . W19 +6.85
CSACK PATH
(i:.) (i:.)
15 3 0.3 .GO059 +6.8911 33 i 36 -,00017 +7. w
lg D 34.3 0:3 ,0CG26 +7.2020 v % 0.3 nmm +7.14
FIG. It lNSTRIJhlENTATION LAYOUT AND CRACK PATI-– TEST 39
00.71.8
?:24.86.3-7.1
;::
:::8,48.58.58.68.47.9
I
I
-14-
in,tervals across the plate in order to obtain strain values, at intervals of one-tenth
the net plate width. Since a double fracture occurred in Test 36, a duplicate test
(Test 39) was conducted.
Rectangular strain rosettes consisting of three SR-4 Type A-7’ strain gages
were used to determine the principal strains at various locations on the surface of
each specimen. Since the component gages were of a finite size, it was obviously
not possible to measure the strain in three directions precisely at a point; however,
since the three component strain gages of the rosettes had a I/A-in. gage length,
it was possible to mount. the three qaqe elements within a O. ~-in. diameter circle,
and this was considered to be satisfactory under the circumstances. In the tests
reported here, two rosette layouts of the component gages were used.. Photographs
and drawings of both gage layouts are presented in I?ig. 1Z. I.n layout A (Tests 33,
34, and 35), the diagonal gage was centered, directly above the horizontal gage,
and the vertical gage was mounted on the side near the initiation edge; in layout B
(Tests 36 through 39), the vertical gage was centered directly above the horizontal
gage, and the diagonal gage was mounted. on the side amTay from the initiation edge.
The gages for dynamic measurements were connected in the customary
Wheat stone bridge circuit. Three similar electrical strain gages, which vwere iso-
lated from the specimen, were used as resistances to complete the bridge circuit.
These bridges were excited by direct current and their outputs fed, to the recording
oscilloscope channels. A diagram of typical. circuits is shown in Fig. 13.
The crack speed was measured with a system of six surface crack. detect-
ors which broke as the fracture traversed the plate; at each detector location in
Tests 36 through 39, two 6-in. detectors (Baldwin SR-4 Type A-9 strain gages)
were wired in series to give a 1Z–in. detector, The breaking of a crack detector
opened an electrical circuit and caused a stepped c’hange in voltage. A diagram
of a typical crack detector circuit is shown in Fig. 14. From a. knowledge of the
spacing between detectors and the elapsed time between successive interruptions
of the circuit, the speed of the fracture could be computed. This system gives
an average surface speed of the fracture since the crack front location is not
known precisely at the time the detector breaks. Thus, all calculated speeds
were rounded off to the nearest 50 fps. Crack detector calibration was obtained
-15-
., .,, .,. ,.. ,, ,, ,.”,, ,,,,.”wm
.-
..<YE; ‘#j* .,.-. . . . . .
..—
< 0,7 in.
Gige hyoui A Gage Layout B
Fig. 12. Gomponent Gage Layouts A and B
L“ 46“
‘l—
‘“ 1
II-Go’
1A 1“-.- -—....
L-$
m ---A . . 0 CRACK IXTECTOR
CALIBRATIONSWITCHES
,“ +=
[ —.l=.-
D-L—
OUTPUT
~“
Fig. 14. Crack Detector Circuits
Fig. 13. Strain Gage Circuits
-16-
TRIGOERINGDEVICES
l+-TRIGGER$IGNAL70 SWEEP GENERATOR
{1\,,,0.,, ,HOMTEO
FOR CALIBRATIONFOIL 1
Fig. 15. Trigger Circuit
by successively opening switches ar-
ranged in series with the various de-
tectors and recclrding the trace steps.
The switch locations are indicated in
the circuit shown in Fig. 14.
A triggering device, referred to
as an external trigger because of its lo-
cation, was uti~!ized in these tests to
activate the sweep and spot intensifying
circuits in the recording equipment.
This trigger consisted of a strip of alumi-
num foil which was broken as the piston
device drove the wedge into the notch;
thus, the circuits in the recording de–
vices were energized just before the
fracture was initiated. To insure that
the circuits would be energized in the event of failure of the external trigger,
a plate surface trigger (A-9 strain gage) was cormec’wd in series with the ex-
ternal trigger. The trigger circuit is shown in Fig. 15.
Static strain gages also were located every 7 in. along the notch line
on both faces of the plate and three sets of back-to-back gages were placed
18 in. above the notch line. However, to simplifiy the instrumentation draw-
ings presented in Figs .5–-l 1, the positions Of the static strain gages are not
shown. Strain readings showed that in each test there was a fairly uniform
static strain distribution across the plate dur
in addition, bending strains were noted to be
tests.
The temperature of the specimen was ~
ng application of the test load;
less than 0.0002 in. /in. in all
:ontinucusly recorded during cool-
ing by means of a ~eed.s and lJorthrup Type G “Speedomax” recorder and copper-
constantan thermocoup!::s located in 1/4-in. deep holes at various points
across the specimen.
-17-
A more complete description of the instrumentation may be found in reports1,2,5
and papers previously issued as a part of this program.
Data Reduction
Reduction of the strain data recorded on 35-mm strip film was facilitated
with a decimal converter and the University of Illinois high-speed digital com-
puter, the ILLIAC.
A brief summary of the data reduction procedure is presented below. The
35-mm film strips were enlarged and the calibration and timing marks scaled on
the enlargements. With the aid of the decimal converter, values of component
gage strain versus time were simultaneously punched on IBM cards, plotted on
an X–Y plotter, and typed in tabular form. The strain-time values were then
transferred from IBM cards to punched paper tape and processed through the
ILLIAC which computed values of the principal strains. The ILLIAG results con-
sisted of tabulated principal strain data, as well as scaled oscilloscope displays
of component gage and print ipal strain traces. The scaled oscilloscope displays
were photographed for later enlargement and processing.
Apparatus and Test Procedure——
In most respects, the apparatus and test procedure used for these tests
was similar to that used in earlier tests made as a part of this program.
The notch-wedge- impact method of fracture initiation was used in these
tests. A closeup view of the 1- 1/8 in. deep notch and the wedge is shown in
Fig. 16. A view of the gas-operated piston device that provides the external im-
pact is shown in Fig. 17. The theoretical output energy of the piston device
was 1200 ft-lb, but calibration tests indicate that the device actually delivered
about 1000 ft-lb.
Crushed dry ice was used to cool the plate specimen. Photographs of
the dry ice containers, as well as a diagram showing the thermocouple locations
for the specimens, are presented in Fig. 18. Typical cooling curves for Tests
33--39 are presented in Fig. 19.
After the instrumentation was mounted on the plate and the plate spec i-
men had been welded to the pull+ heads, the specimen was stressed at room
-18-
Fig. 16. Closeup of Notch and Tipof Wedge.
temperature to the test load to check
Fig. 17’. Piston Device Used forFracture Initiation.
the behavior of the strain gages; at the
same time, it was possible to ascertain the static strain distribution in the
specimen and to obtain the test load strain values for each dynamic gage. Any
gages that displayed faulty or questionable response were replaced and the
newly installed gages were checked by another test load cycle.
In Tests 36-- 39, wiring between gages and oscilloscopes was double
checked by shunting each gage with a resistor and noiing the direction of trace
movement on the face of its respective oscilloscope. This gave a positive
identification of the scope trace corre spending to each gage and the direction
of the compression calibration for each gage.
At the time of test, the cooling tanks were filled with crushed dry ice;
when the test temperature was approached, the plate was stressed to the test
load. After loading, the recording equipment was calibrated, and the gas oP-
erated piston device pressurized. When the desired specimen temperature was
obtained, the recording camer~s were started, and tlm gas–operated piston de–
vice was fired to drive the wedge into the edge notch to initiate the brittle frac-
ture.
ANALYSIS OF TEST RESULTS
General
The re suits p~resented here are based on all of the applicable data ob-
-19-
Thermocouple Locations
@-. ‘~:$,
$ ‘* ,.,
R
“b,;$,...,,”
,1
Individual Cooling Tank
Fig. 18. Thermocouple Locations
.20~ I I I I I20
J40 60 00 100 120
60
TEST 3440 \
. .
20b,Y
0.4 _
-20020 40 60 eo 100 120
80
\TEST 3S
60
-20 1 1 1 1 1 1 I0 40 80 120 160 200 240
TIME — min.
Fig. 19. Average Cooling Curve s--Tests 33--39.
Cooling Tanks Mounted Prior toTesting
and Cooling Apparatus
‘Fl=T=--20 I I I I I I
60
40TEST 37
\
20 \ . ..—. ,“. — .——:
0 — ,.-—,... . ..—. .—— — --“! “---’”–’
-20
60
40
TEST 39
00\
20 . . .. . ,—-.. -——
:
0 I=-. ~
-20020
T[h?E — ;in. ‘0 ‘0” ’20
-20.
II I I I
0bBz7~o 05 10 15 0 05 10 1!
FIG 20 TYPICAL STRAIN TRACES AND COMPUTED PRINCIPAL STRAINS
FOR ROSETTES LOGATED AT VARIOUS DISTANCES FROM
THE FRACTURE
tained as a part of this program. The
extensive strain data obtained in Tests
33--39 are pressnted in this report for
the first time. ‘The use of strain rosettes
in the latter tests permits the plotting of
contours of maximum principal strain, as
well as contours of vertical strain, on
the surface of the plate as the fracture
propagates acress the plate. The data1,2
obtained in earlier tests support the
results of the more recent tests which
are presented here.
Recorded Test Data— —
Typical strain traces for com-
ponent gages located at various dis-
tances from the fracture may be seen in
Fig. 20; the three rosettes are from one plate and are mounted directly above
one another at distances of 1.1 in. , 4.1 in. , and 7.1 in. from the fracture.
The strain traces shown are representative of the strain-time curves obtained
from all tests conducted as a part of this program* The decrease in maximum
strain magnitude and the change in pulse shape as the distance between the
gage and fracture changes may be clearly noted.
The vertical and diagonal gage traces exhibit a general shape character-
ized by a fairly steady and rapid increase in strain to a maximum (peak) strain
value as the fracture propagates past the gage; the peak strain is followed by a
decrease to a strain level associated with the removal of external load. For
gages located close to the fracture, the peak is very sharp; for gages located
away from the fracture, the peak is of a lower magnitude and the pulse extends
over a longer time.
-21-
.0040STRAIN TRACES FROM
COMPONENT G?4GE5
0030 —.. .-
i00?0
<= i
I \
I 0010 — .-.,. - —— -.. p
l..z r /;:~ ,.O ,9-!
–––-+ )r\~.,\—
---_-. —0 ------ .— - -.’—-/- ~ . . .. J--.-— —-—?,-” d. \.. _..A. ~-
–0010
–0020005 1,0 15 ?0
— ,..PRINCIPAL STRAINSAND DIRECTION
I OF G). ~~x I I I
t----t ~
G.*__dil” b+’”
R’W+1 1 1 I0 05 10 (5 2.00
TIME — ml(ll,cc”ndx TIME — m,ll,,e,..d.
FIG 21 STRAIN TRACES FROM GOMPONENT GAGES NosAND
I , 2 ~:~r 33;PRINCIPAL STRAINS FROM RoSETTE A —
The horizontal gage traces are characterized by three major changes in
strain. The strain trace first exhibits an initial relaxation of compressive
strain, and this is followed by a compression Pulse corresponding to the ten-
sion peak of the vertical gage. Finally, the trace exhibits another relaxation
of compressive strain before leveling off at the final strain value.
The strain-time curves of the component gages obtained from Tests
33--39 are presented in Figs. 21--8!3. AU strain traces are plotted such that
the strain at zero time is the initial test load strain.
The second peak occurring in the component gage traces (and principal
strain traces) in Figs. 43, 45, 47, and 50 was attributed to electrical effects
associated with the recording equipment (although there is still some ques-
tion regarding this matter), and not the double fracture of the plate specimen
in Test 36.
The crack speed detectors were used to measure the approximate surface
fracture speed and to aid in determining the location of the surface fracture at
any time during the brittle fracture test. The average surface speeds of propa-
gation of the brittle fracture for Tests 33--39 are presented in Table 4. It will
(Text continued on page 39)
-22-
TABLE 4. FRACTURE SPEEDS
All distances are measured along the crack path.Speeds are rounded off to nearest 50 fps.
DistanceBreaking Time
II st anteDetector between Speed Detector between
Breaking Time
Time Interval Time IntervalSpeed
Detectors Detectors
(in,) (milliseconds) (fps) (in.) (milliseconds) (fps)
TEST 33——
A 0.7612.02 0.36 2800
B 1.12
13.16 0.37 2950c 1.49
6.02 0.22 2300D 1.71
6.05 0.25 2000E 1.96
TEST 34. Fracture Passed above Crack Detectors
A
B
c
D
E
A
B
c
D
E
F
8.00
8.00
8.00
8.00
7.10
6.64
7,32
7.15
14.47
TEST 35——
0.770,23
1.000,14
1.140.24
1.380.20
1.58
TEST 36
0.310.13
0.440.15
0.590,23
0.820.20
1.020.52
1.54
2900
4750
2800
3350
3500
4150
3050
3000
2800
A
B
c
D
E
F
A
B
G
D
E
F
A
B
G
D
E
F
7.02
7.10
6.50
6,97
13,60
7.14
7.16
6.71
6.94
13.34
6,73
7.18
7.18
7.24
14.0
TEST 37
0.96
1.10
1.28
1.50
1.78
2.40
TEST 38——
0.34
0.49
0.65
0.85
1.07
--
TEST 39——
0.06
0.25
-—
0.69
0.95
0.14
0.18
0.22
0.2s
0.62
0.15
0.16
0.20
0.22
-—
0.19
0,44
0.26
0.461.41
4200
3300
2450
2100
1850
3950
3750
2800
2650
2950
2700
--
2300
2550
FIG 22 ~NAIN TRACES FROM COMPONENT GAGES !40s 4, 5, AND 6,PRINGIPAL STRAINS FRO?,! ROSETTE B — TEST 33
00’0
Q030
0020
<E
1 001010-”z
a:.
;lgo_, __,
0N-” .-~
1
I
-0010 —–— ——
-Oozoo05
*15
T,ME — ., !N,ecomd,
,. IF,:IPAL STRAINSAND DIRECTION
OFGm. ~~x
9— —. —.- _~
+“m
k%.” ___,,
To 05
T,,.!E — ,,, ,, second,
FIG 24 STRAIN TRACES rROM COMPONENT GAGES NOS 10, 11, AND 12,M. D PRINCIPAL ST RALNS FROM ROSETTE D — TEST 33
——,30,0 10G
STRAIN TRACES FROM PR,NCIPW ST17AINSCO&APONENT GAGES .KD DIRECTION
OF %m ~~x
0030 :!50 :
:.
I
0020 120 :<
< :j :
1 0010 ~ IIo—.~- -.
90 :7–” !,-. ,
z ~ :
:,-, ___ ‘
d ~ ~:
d, ,---
0__ L-!___ -____ .-J-,---- ?..’ ‘
‘. z 60 ~_ :.,<__ ______ ,= —--- .+.l, - - f
b
5
–0010 30 ~E6
–0 02000,5 1,0 1,5 200 05 10 !5 200
TIME — r.,lti%ecomd, TIME — m,!h,,cond,FIG 23 STRAIN TRACES FROM COMPONENT GAGES NOS. 7, 8, AND 9,
A MD PRINCIP&L ST R&l NS FROM ROSETTE C — TEST 33
IixLQI
lIME — m,ll,s,co.,d, TIME — mill,cco.d.
FIG 25 STRAIN TRACES FROM GOI,i POMENT GAGES Nos 16, 17, AND IB,AND PRINCIPAL STRJINS FRo PM ROSETTE F — TEST 33
TIME — m,,,l,,,,.nd,
.FIG 26 STRAIM TRACES FROM COMPONENT GAGES NOS 25, 26, LNO 27,
AND PR!NGIPAL STRLINS FROM ROSETTE I TEST 33
c
E_<-_”_
-00>0
–OO?OO
STR&W TRaCES FnomCOMFWMENT GhGES
1
I10 15 2C
PRINCIPtiL STRA,hWAf4D @l!?E. TlOtJ
I
---,--
------
FIG 28 STRAW TRACES FROM COMPONENT GAGES NO’S lb 2, AND 3,
AND PRINCIPAL STRAINS FROM RoSETTE A — TEST 34
I -0020.-0! 015 )0 1!5 ‘ ,1: II T!m — m,l,,,,cmf, TIME — m:ltisecmnds
FIG, 27 STF+AIN TRACES FROM cOMPONENT GAGES NOS 21S 29, aND 30,AND PRINGl PAL STRAINS FROM ROSETTE J — TEST 33 I
m 130
s7Rw TRnCES FROM PRI:4CIPAL STRAINS
COMF.3MENT GAGES LXD DIUFCFIO:4 I
0030Fww?+H!
‘“”’mm:!:;~~r~[i“ 05 10 15 20 0 10-.
TIME — mll,aecond$ TIME — m,ltiae:omdi
FIG 29 STRAIN TRACES FR Oh! COMPONENT GAGES 140s 4, 5, ANO 6,
ANO PRIPIG IPAL STRAINS FROM ROSETTE B — TEST 34
I—:2200 I I I I I I I05 10 15 200 05 10 15 2 O*
TIME — m,ll,s,cm,, TI,JE — m,lfiwc.,,d,
FIG 30 ST17JLIM TR,LCES FROM COMPONENT GAGES Nos 7, B, AND 9,AND PKi MGIPAL STR &l NS FSOM ROSETTE C — TEST 34
TI,ME — m!I,, cco,d, T(ME — ,,lti, e.o, d,
FIG 32 STRAIN TRACES FROM G0h4P0NENT GIIGES NOS 13, 14, AND 15,AND PRINCIPAL STRAINS FROM ROSETTE E TEST 34
.. . .STRAI’I TR4CES FROM
COI>FO!4E?JTGS.GE5
,0030
0020
+s 3
I ooro ,,0., [. /
<E. -u- _~J-’” “::
0 ,,-” ,,
-0010
–0020005 so 15 20
TIME — mlll,,,mmd, TIME — m,ll,,conds
FIG 31 ~N~lN TRACES FRoM GOMPONENT GAGES tiOS 10, 11, AND 12,PRINCIPAL STRAINS F ROM ROSETTE D — TEST 34
STRAIN 7RACES mov ,WK,F*L STRAIN*CGUPOME121 GAGES AND OIREC.10,4
OF %. ~~x ~
0030 k50
I
0020 ,?0
<*
—... /“
1 0010 n_O ——. —--- -
. ,6-” .,”rr< —E \. I ,-0— — — —-—. - -J \
0- .—L _,,-” -=/ ~\ ,.-.
,.‘.< /
-0010 30
–D020D05 10 !,5 200 05 10 15 200
TIME — nlll ,,<”,5, 7,ME — m,lti,,,cnd,FIG 33 ~N&N TRAcES FM OM COMPONEUT G.4GES NOS 16, 17, ANO 10,
PRiNGIPAL STRAINS FRIIM Rn<!=TTF s — TCCT X.
Ir-dU-I
rRA’t? TR&CES 6301.1CO1.WQtfEf4TchG:S
.1.
15 2(
PRlt<cIPAL sTRAINS !M3 O,RECTlON
OF S... ~~x
/’J’
—_ —--
1 180
,---
fl !
T- “60.J i,\
L- --
30
0, 1“ 15 200
T,WF — m,,l,,. co, d:
FIG 34 STRAIN TRACES FMOM COMPONENT GAGES NOS 22, 23, ANO 24,
IIN D PRINCIPAL STRAINS FROW. ROSETTE H — TEST 34
I I
%
\Ii, ;J I ,::,.,(-.
,i
\‘ . ---
-— —-..
:&0 05 10 15
PRIIW+AL sm,,,~sAME c,RECTIO?J
OF Gm ~~x
+–-—
*
Pi,,
E+---- _i\
. ..1 i\
cm,”_—— —— 1 ,.&\\\\/---
,Y~__
1’
J_L_05 >0
““. . STRfdf4 TRACES FROM ,, INcIPAL 5TRLIPJS
CCJ4WMEMT GAGES &vO OIRECTION [.,0, %2, ~~x
0030
,0020
<~
1 .2010 ./15-,?
:
0___2>-x —-.—// \
~1 ,--”
?
–001 0 g.
–OO?OO05 ,0 15 200 05 la 15 2 O*
lIME — m,ll,,comit TIME — mi~,,condf
FIG 35 STRAIN TRIIGES FROM COMPONENT GAGES NOS 25, 26, AND 27,
A MD PRINCIPAL STRIMNS FROM ROSETTE [— TEST 34
““. .
STRAW TRACES FRDV PRIVCIPLL STRAINSCC+JPOflEtlT GALES AND DIREcTlOM
OF %,. ~~x
:0030 150 :
:.
I
002Q:
,20:
< 2~
i ?,.?,/
z
eo
~ Go,o/ \,
30 ;;—
z j i,~___ ‘/ \ :
,,.”a ,< , ‘== \
E5
. ,3=. _,
i’
\ f_ ,— -——.i
~60 $0—<L-l——— .. ——-—-. ---__, -J’\J-\.-/. ‘“’” ,2 -\J’\, !J .
, -----
‘\ ‘ ‘;!,/
–0010 30 ~.6
–0020005 :0 1,5 200 05 10 15 20°
lIME — m,ll,K.”dt lIFVE — . 18,,,.”4sTI,.!E — miltis,mnd.
FIG 36 STFIAIN TRACES FROM COMPONENT GAGES !40S 22, 23, AND 2+,
AND PRINGIPllL ST R.4!NS FROM MOSETTE J — TEST 34
C,,,40
! 5TRhl&4 1R>ICE5 FROI3040
0030
R
~-----
..——.
—
;TRAIN TRACES FROMcOMPONENT GhGES
I
.al,4cl PAL SrRl !:sAND OIIIECTIONCAGES
----
. .. . .
—
e
0020
<s
I,0010“k.-,.aEm s– ,
0
5–,
–0010
–OC200[
—.-—-
—
,--
/’-
/- E+
... (
.+, -—— __
——,—
L+-----
i200
TIMZ — m :0,econd,
FIG 3e ~q&N TRAcES FROM GOh!PONEPIT GAGES NOS 1, 2, AND 3,PRING! PAL ST Rfll NS !=Bo M ROSETTE A — TEST 35
TI,;,E — m,l,,,ccndi r,9E — m,lll,,<ond, TIME — m<llseco,.ds
FIG 39 STRAIN TRACES FRO!.! cOMPONENT GAGES NOS. 4, 5, AND 6,A MD PBINGIPAL ST R,MMS FROM ROSETTE B — TEST 35
PRINCIPAL sTSAV45J,W GIRECTICN
II
.. .f .7s.,r4 TRicmmm
COMPONENT GAGES
0030
co ?9
<s
1 0010 ‘0-’ /(L ‘“1. y,
a //1E __,2:, _ __‘1 “ ,y/J~L_
0 ,,-” /’” ~-..r–––l. —__..-
-0010
-0020005 10 15 ?0
,003.
3020 ———
031C 7-”
,-,
c,-”——————
-Oclo
–o@zoo[
GILGE5
._.---,
-—---+ .=,.-———- -—--- .
I
i10
TIVF — m,l,,,,:md, W4E — mil,,,,c,,d, 14!!E — m!!,,,,. ai,
FIG +1 STRAIN TRACES FROM GOMPONENT GAGES MOS 10, II, AN!3 12,AND PRINGI PAL STRAINS FROM ROSETTE O — TEST 35
FIG 40 STR&lN TRACES FROM GO1.4POPIE:4T GAGES NOS 7, 8, AND 9,A NO PRINGI P&L ST RLIMS FROM ROSETTE C — TEST 35
-0010 L1--OO?OO
FIG 4?
0040r0030
0020L
–0010r–00200
,.005’415rRhlr4 TRhCES FROk
CGWWME!+T G&GES
!
I
1
j /
Y !
;1I
I
I
I )
I
-, . . _.-— — —- ---
L10 15
TIME — mll,l,momd,
STRAIN TRACES FROM COMPONENT GAGES NOS. I 2 , AND 3,
A MD PRINCIPAL STRAINS FROM RoSETTE I — TEST 36
STRAIN lRfiCES FROMI Co,,Pot
T$i/—./..,-{,I 10
IPRI:4CIPGLSTRA,&W.4N0 OIRECTIOS
OFcm. ~~x
.—l.
i’,,50 :
:.
m,20 z
.
1FIG 44 ST17fi[N TRACES FROM COMPONENT GAGES t+OS 11, 12, AND 13,
AND PRIN Cl PAL STRAINS FHOM ROSErTE 3 — TEST 36-_
TIME — 11111,,cond, TIME — m,lllae.ond$
FIG 43 STRAIN TRACES FROM WMPONENT GAGES !40S. 6 , 7 , AND 8 ,
AND PRINCIPAL STRAINS FROM ROSETTE 2 — TEST 36
Ir-dm
TIME — mill,eca”d! TIME — m,lll,,co.d,
FIG 45 STRAIN TRACES FROM COMPONENT GAGES NOS. 16, 17, fiNO 18,AND PRINCIPAL STRAINS FROM ROSETTE 4 — TEST 36
FIG 46 5TRLIN TRJWES Fn OM COMPONENT GAGES ?40s ZI, 22, AN Dz3,AND PRINGb PAL STRAINS FROM ROSETTE 5 — TEST 36
TIWE — m lh,,co,d, Ti)AE — m,ll:e:omds
FIG 48 STRAIN TRACES Ft701M GO1.iPONENT GAGES NOS 26, 27, fi14D 20,A Ma PRIMGIPAL STR.4!NS FROM ROSETTE 7 — TEST 36
s7nw4 7RficEs FnarnC0Wt3NENT GAG=
0030
,0020
+5
1 ,0010 i
. ,-,<2. 4 . . /_, -,
0–-/,4-H
–0010
–00200D5 10 1,5 20
TU4E — mlll,l,cond, TIME — m,,ti,,cmnd,
FIG 47 STRAIN TRACES FROM GOMPONENT GAGES NOS 24, 4 , ANO 5,AND PRINCIPAL STRAINS FROM ROSETTE 6 — TEST 36
TltAE — m,r,, cm. d, TIIHE — m,lti,em”d,
FIG 49 STRAIN TRACES FROM COMPONEMT GAGES NOS. 29, 32, AM031,AMD PRINCI PAL ST RALNS FRO M ROSETTE R — TEST 7G
11F4E — v,lti.ec.ndm TIME — milti,,conk,
FIG 50 STRAIN TRACES FROM COMPONENT GAGES MOS. 25, 9, ANO 10,A No PR!MGIPAL ST RAi MS FROM ROSETTE 9 — TEST 36
0030
0020k
0
F-._.-J-
,,–”
-0010
FIG 52
TIME — m,lll,ec6-”d.
:
H FROWr GfiGES
._— -.
,> -=
J20
,,%,—
..,..——4-
STRAIN TRACES FROM cOMPONENT GAGES NOS 33 , 19, lkN020,
AND PRIMGIPAL ST R&l MS FROM ROSETTE II — TEST 36
-“ ,.
STRAI,Y TRdCES FROM PRWCIP3L” STFIAI?45
COMFONEtiT GhGES aHO OIRECTIOll
OF Emm ~~x
150 c
~
,0030I1. %, \
.
I,$,
j \\g
0020 I } \120
;~:
i i
<</
&
AI I i, --, :
aIj
1 0010,,-” –.;; ><?
90 :
1~~ il jl “‘m“~ =: 5L ,--- i I - E/
0/ >,J? , I ,0 z
,0-” ----- . .;.— -’ -\-/- . -.—. $
:
~
-0010 30 ~.E
-0020005 10 15 200 05 10 15 20°
TIME — mill,,,.omdb TI.WE — milm, cb.d,
FIG 51 STR&lN TRAcES FROM COMPONENT GAGES NOS 30, 14 &~~6,AND PRINCIPAL STRAINS FROM ROSETTE 10 —
0030
r
F,-.
-6010
...—
rn>ifi TRACES FROMCCW=OKENT GflGES
+
F.,,-‘f I
d,= - –-----
TrlME — m ,emmdb
PRINCIPAL SIRAINSh,,. DIRECTION
>,.———
I
7:-.=_60
.
&
FIG 53 STRAIN TRACES FROM COMPONENT GAGES NOS. I , 2 , hNO 3 ,AND PRINCl PAL STRAINS FROM ROSETTE I — TEST 37
woI
0040 k; STR,?IN TR>CES FROM
COMWJ”E%T EhGES
j
0033
0020
0010 A
w-” /
.’ ‘!
0
E-H
0010
–002000,5 10 15
TIME — mlllme<mdm
1 moPRI!XCIPLLSTRA!!JS
AND DlRECT10P4 1
OF c... ~~x ! I ,50
,~1,li
,20
I;p /
I ii ‘.,,,’”
e Ow ,— — — — 90
J●.6,
1. —- ____ ——_ , ~.
_/ ‘,—— ———,●mi, ~
3G
05 1.0 15 28
TIME — mill.,eoand,
FIG 54 STRAIN TRfiCES FROM COMPONENT GAGES MOS 6 , 7 ANO S ,
btiD PRINGIPAL STRAINS FROM ROSETTE 2 — ‘TEST 37
0040STRAW TMCE5 FROIA
CO!APONE?IT GAGES
GO30
,0020
<<
[ 001,
~ ~
: /-.—
k / .._= .——<,<-,
0 ~-- -— —-—.,
1,-”
–0010
–OO?OO0,5 10 I ,5
TI},IE — mll[,,,,o.d, TIIAE — vl?,,cor,d,
FIG 56 STRAIN TRACES FROM GOM PONENT GAGES !40S 16, 17, ANO 18,ANO PR!NGI PAL STRAINS FROM ROSETTE 4— TEST 37
‘“”’orrrrl i ‘0iI TIWE — ..111,,.0 ”{, TIME — miltiaec.mds I
FIG 55 ~H~ TRAcES FROM COMPONENT GAGES NOS. II, 12, AND13,PRINGIPAL ST RJ!NS FROM ROSETTE 3 — TEST 37
FIG 57 sTRAIN TRACES FROM GOMPONENT GAGES f40s 21,22 , AMD23,AND PRINGIPAL STRAINS FROM ROSETTE 5 — TEST 37
I c,..6=P — SIXV lVMIOZ l&OH wow ‘“W3 ,0 NO I1ZIU1O
W .1 — NIVMLS
,,,, b,o — Slxv lVLNOZIMOH Now ~~3 do Nollmwo
.
,U!,UI — NIWWS
,...!/ .1 — NIWIS
X,* 0 ,~:
STRAIN TRACES FEO,$I PRI,MCIP*L sT. &lhsCC,VPONE14T GAGE5 AND DIRECTION
OF %. ~~x
0030:
150 :
:m
r
!I
00?0g
120:
+!
i=
_e~; 2
I 0010.— _ _ __, \] :
![90 ~
~ ,,-, ●m,t :
z 1
J
5G \
-. L+-_x _ ,— ——— i Ii E
0 .— _-30-,
._-..fJ~ _ >&__
‘1
.––f J ;~-T— x:,+ 60 “~,(J7 ,
&
i ~!\
1:5
-0010
I ‘D i
–o 02000,5 I ,0 1,5 2,0 0 05 !0 1,5 2.00
TIME — mill,Womi, TIME — .,:w,*co.4s
FIG, 62 STRAIN TRACES FROM GOMPONEMT GAGES NOS,30, 14, AN D15,A MD PRIMGI PAL STRAINS FBOM ROSETTE 10 — TEST 37
TIIWE — mlll,,,cond, TIME — mill,,,cond,
STRM/ TRfiCES F.W?4
CCH4FQNENT GAGES
,0030
0020
<<
,1
~ 0010 {
. 20-” -a
/ / ,; ‘\E ,.-,..3 —. -,, -
_-. —-. —, --- ~= - .
0 ,,-.r.
-— —__ —- —___ /’ ‘J
-0040
-0020005 1.0 1,5 20
TIME — milll,,cnnd,
FIG 63 ~J#N TRACES FROM COMPONENT GAGES NOS. 33, I 9, AND 20,PRINCIPAL STRAINS FRO M ROSETTE II — TEST 37
FIG 65 ST fiAIN TRhCES FROM GOMPONENT GAGES NOS 6, 7, f4Nll 8,ANO PRINCIPAL STBAINS FRO M ROSETTE 2 — TEST 30 I
,0040 I S7R<4 TROCES FROM PRINCIPhL STRAIMS 1~
I -00200I
05 10 1,5 200 05, 10 15 2.00
T,,,E — mil,l,, co.i, TIME — m,lti,,cond%. IFIG 66 STRAIN TRACES FROM G3MPONENT GAGES NOs 11, 12 #TD ~8,
AND PRINCIPAL STRAINS FROM ROSETTE 3 —
. . . .I I I’T%LR:GRT1I‘R’’””L5“*’”’I I I IANODIEECTION
‘“3”WFHM”:
–D0200 I I I 1 -805 10 15 200 \
T,,,E — n,.lll,, m”~s TIME — m!lhKmd G I
0040 760
STRAIN TRACES FRO% PRIXCIPAL STRAINSCCUPOHENT GfiGES *MD OREcllON 1
TIME — m.lliecomd, Tl!dE — mlfi,,c.”ds
FIG 67 STRAIN TRACES FROM COMPONENT GAGES NOS. 16 ,ANO
17, AND !8,FRIN Cl PAL ST flAINS F ROM ROSETTE 4 — TEST 38
0040
0030
0020
+<
] 0010
~
:L
0
-0010
-0020
STR?MM TRiCES FCOMFilllEliT G~{
B
,--- ,-- \,-” /.// \
,-.-. ,,-.,-—–-K. .--=---’-- ,,.k,-”
‘k.
05 10 15TIME — .,,ll,,,.mt,
FIG 69 STRMN TRACES FROM COMPONENT GAGES NOS 24, 4, AND 5,AND PR!NCI PAL STRAINS F ROM ROSETTE 6 — TEST 38
TIME — n,lti,,.end<
1 1
I
OOGOsrnm 7R2CES mm
CC.WPONCNT GhGES
OQ30 I
!!0020
<:
I 0010 ;
~m-”
:
k~
..7
~ A ~ -— ——-.~-
-\_-,6-, ‘.//
-0010
–002000,5 1,0 15 Z@
TIME -- mll,l,, cond, Tlt#.E — m,ll,,,cood,
FIG, 70 STRAIN TRACES FROM COMPONENT GAGES NOS. 26, 27, ANO 28,Mio PRIN:I PAL STRAINS FROh% RoSETTE 7 — TEST 38
PRIPICIPAL SIRh’!45
.,.3 DIRECTION
OF~m,. ~~x
150 :
:.
I
.!10 z
f-.,’
. . . .~
..,!
r6
0 05 10 15 ? 0°
TIME — .m:l,s,cor,db liME — Wl,,s,,o”d,
FIG, 72 STRAIN TRACES FROM GoMPONENT GAGES Mos 25, 9, AND 10,AN!l PRINr.l PA1 STRAINS FROM ROSETTE 9 — TEST 36
-oozo~ —’ 10,5 10 15 200 05 1.0 15 20°
T114E — mlll,,ecmd, TIME — mill.s,wnd,
FIG, 7i STRAIM TRACES FROM COMPONENT GAGES NOS. 29, 32, AND 31,AND PRINCIPAL STRAINS FROM .~OSETTE B — TEST 3B
IwmI
,SU6. P — S,XV 7vLNOZI !dOH WOW ‘% 40NO1133H(0
,
,.!/ “1 — Nlvnls
-1-..![.1 — NIVtlLS
—.
I \ II
I0030
k
–0010
r
—
—
—
L
i—
—
,TRMV TR2CES FRC\~COMFOM
.._._=
TIME — m,ltisemr.d.
Iz 0°
FIG, 7S STRAW TRAGES FROM COMPONENT GAGES ?40S 16,17, AN016,
AND PRIMGIPAL STRAINS FROM ROSETTE 4 — TEST 39
0040 Ima
S1flbW TR&CES FR018 PRINCIP&L 5TRfilN$
C4MF.3ME12T GfiGES AKD DIEECTjON
OF km. ~~x ~
0030 ‘1 150 ~!? s
II.
il I0920
j L- 120 ;
I “.
. ---< :
+ i. >
I 0010 fl:
).?
90 :
z :
a ~.
,mn,
E ,-, , ~z
m E
/ _-— ——————
024-” ,,
-’ 60 ~
--./- %% __,’ ;/ ‘-J ‘-’ .
Y :
$
–0010 30 ~
.6
-0020005 3,0 1.5 2,0 0 05 1.0 15 20°
TIME — mlllt, tcm~s TIME — rnz%,cmnds
FIG, 00 STRAIN TRACES FROM GOMPOMENT GAGES MOS 24 , 4 , AND 5,AND PRING!PAL ST RAINS FROM ROSETTE 6 — TEST 39
. ..”
STRAIN TR6CES FRO!dCOMPONEflT GAGES
,0030
0020
<g
I .Ooro 1 /
~ , ,.” / ‘.
: ) Ik ,,.. .- _- - . —.;—- ,.,~ ~~,- ,
0 ,,. ” .- -- ~J ~’-,—-—- “‘“T - —._.
-.. .—
–0010
-002000,5 1.0 1.5 20
..PR,MCZPAL STRAINSArdD OIRECTIOM
OF %m ~~x
150
,
}120
J\
f- [ \ -.i, ; \“
0 ,/’—-,
1 90., , Ii
~.- .— ~
=.,, l!
,/ - .,1(
/,60
-. -.,, .,”
‘( +-..,
. . ..-. .
30
0 05 10 15 2}
T4ME — m,ll,$ecom~, TIME — m,lti,econdsFIG 79 STR.4[N TRAcES FROM COMPONENT
ANDGAGES NoS. 2 I , 22, AND 23,
PMbNGIPfiL STRAINS FRO M MOSETTE 5 — TEST 39
.“. ”srRhlN 7RficEs FROM
COMPONENT GAGES
.0030
0020
.:s
I 0010 f
): ‘1z 2 @-v / \n
: . ~ ‘Lk 2:-D, “ P.J 1
,4 . _——0
,: +.’ ‘+
-0010
-0020005 10 1,5 2,0
TrWE — m,,lll’,cemd>FIG. 81 STRAIN TRACES FBOM COMPONENT GAGES NOS,26, 27, ANWZE,
ANO PRhNCIPAL STRAINS FROM ROSETTE ? — TEST 39
0030
0020
L
0 F-/.’,9-”
FIG 02
RA13 TR
Cowmn
—
rlME — milll, mmds
STRAIN TRACES FROM COMPONENT GAGES MOS 29,32, AN D31,A No PMINCIPAL STRMNS FROM ROSETTE e — TEST 39
TIME — milll,ecnnd, TIME — milti,,con,di
FIG B4 STRM4 TRACES FROM COMPONENT GAGES NOS 30,14, AND 15,
A NO PRINCIPAL ST RAi MS FRO M ROSETTE I 0 — TEST 39
0040 100
STRAIN TBACES FRO.W PRIMCIPAL SIR&l!M
CC?+ POHENT GfiGES ANO DIRZCTION
OF Em.. ~~x
~,
0030 150 ~
s.
I
0020g
,20:
< i
~ z
Jle ,- .
I 0010,’ :
—.,’ -90 ~
\. -, _ /,
z m-, .mo, / :
‘\: I ,, i,
/ z
k —,-> — —- f 4 :._ i,,, ,/ ] —
0
-2T-c- Yv{”-- -~:-. 1– ::- –-’ - ‘, 1
69 ~
●“ah,_ L- ——.. _____ .W
b~ *
6
–0010 30 ;
gm
–0020005 10 15 200 05 10 15 20°
TltiE — m,lfi,,cmd, TIME — m,lti,em”d%
FIG 83 STRAIN TRACES FROM COMPONENT GAGES NOS 25, 9 , AND1O ,ANO PHbNCIPAL STRAINS FROh\ ROSETTE 9 — TEST 39
ILdcoI
“... . .
STRAINTRACESFROM P31NUPAL 5TRAINsCOUPONENT GAGES dND OIRECTION
or %. ~~x
.
FIG 85 STRAIN TRACES FROM COMPONEMT GAGES NOS 33, 19, AND20,AND PRINGt PAL STRAINS FROM ROSETTE 11 — TEST 39
1 II
—
-.
.-.
..-
—
-39-
be noted that these speeds ranged from 1850 to 4750 fps, and are thus in the same1,2
range as those reported for earlier tests.
Computed Print ipal Strains-—
To compute the principal strains from the rectangular strain rosette equa-
tions, it was necessary to determine strain values from the three component gage
traces at selected times; the times were selected arbitrarily at points correspond-
ing to changes in strain in the component gage traces.
The rectangular strain rosette equations used in computing the principal
strains and the direction of the maximum principal strain are as follows:
r #- .1
E
[
lEmin . —2
v+ ‘h- 2 (Eh - Ed)+ 2(Ed – &
1-1
-1.2Ed-’h -’vQ = ~ tan
2
where: Emax =
0=
E =v
‘d ‘
‘h =
J
maximum print ipal strain
minimum principal strain
angle between ~ and the positive X-axismax
vertical (Y) component gage strain
diagonal component gage strain
horizontal (X) component gage strain
-40-
The. re suiting curves of the principal strains (~ ~ax and ~~in) and the direction
of c ~ax with respect to the positive x-axis {6] for Tests 33--39 are presented
in Figs. 21--85.
There are several important details that should be noted with respect
to the time— alignment of the component strain traces for any particular rosette
gage. In the case of strain records obtained with gage layout A (Fig. 12), the
vertical gage peaked before or at the same time as the horizontal and diagonal
gages for about 80 per cent of the rosettes. A detailed study of the component
gage traces and the principal strain computations indicated that changes in the
vertical and horizontal strain values have the greatest effect on the principal
strain magnitudes; thus in Tests 36–- 39 it was decided, to center the vertical
and horizontal gages above one another in the manner shown in layout B (Fig. 12].
For records obtained with gage layout B, the vertical and horizontal gages peaked
before or at the same time as the diagonal gages in the case of about 85 per cent
of the rosettes. Typical conmonent gage traces from both layouts are shown in ‘ -
Fig. 86; the offset in peaking times may be noted clearly.
It was believed initially that a refinement in the results could be ob-
tained by shifting the recorded component gage traces to a position where the
maximum strain values would occur at the same time. However, later studies
showed that there was only a small change in the principal, strains as a re suit of
this shifting of the component gage traces. This is illustrated in Fig. 87, which
shows the print ipal strain traces for a typical rosette computed 1) for the com-
ponent strain traces as–recorded and. 2) for the component strain traces shifted
to make the peaks occur at the same time. It will be noted that shifting the
trace does not change the shape or magnitude of the maximum print ipal strain
trace markedly. For most rosettes it would be necessary to shift the trace by
less than O. 1 milliseconds to make the maximum strain values occur at the
same time. This time difference is of the same order of magnitude as the inherent
time error that accompa.nies matching of the trace times during a reduction of the
original 35-mm strip-film record. It should be noted that many other factors also
tend to affect the strain traces, such as the discontinuous nature of the fracture,—
the deviation of the crack path from the notch line, and inherent variations in in–
strumentation. All of these factors and studies tend to justify the decision not
.
-41-
--
.-
0,0020 0.0020
L4Y0 T A
I TEST 34 .sROSETTE A .
i 0,0010.
/
- 0.0010
; l—v.—
I zz z—o — ,—..1” z,.—.”—,.z + :: 0
I3-H -./ ~. l/---- .-.. _
o
‘~-. .-—-—
-0,0010 ~I I I I I
o 0.5 1.0 1.5 2.0 2,5-0.0010
(
TIME — rmllis;conds
““”’;’-1 1 1 1 I
0,5 1,0 1,5 2,0 2.5
TIME — r!,ll,swonds
Fig. 86. Typical Strain Traces from Component Gage Layouts A and
mm -
B!ilB,,,,, .,,,,,., ,,.,[,
m.,rTT! ,, -,,,, 37)
0085 . .A
00 0 .,- .-.i ~,
0003 ‘-’” —-—. , -.
-.= :;- \ ,Lo
-.-. ,. <.——.- —---- ,,-. —... ‘i,{
-0003 .—
Fig. 87. Effect of
,,., - .),,,.. <,”,.
Shifting Component
“l_—~
,,0
EEP!i/-.-—,.,,:i.cx
.x “-”’’’””’”‘“ ““’,,~~.> “.
73
~,
m -., —--— ,,” “,,,— ,,,, ” ‘/
+5o ., 1, ,$ ,0
,,., — .,,,,. ?,.”,,
B.
Gage Traces on Print ipal Strains.
to shift the traces; therefore, all of the priric ipal strains have been computed
from the as-recorded component gage traces.
Discussion of Strain Traces——In general, the maximum principal strain trace for each rosette is of es-
.—
.
—-
-.
sentially the same shape as the vertical component gage trace. The magnitudes
of the peak strain for the vertical gage trace and for the print ipal strain trace
are also nearly equal for any rosette. This is shown in Fig. 88, in which a
comparison is given of the maximum principal and vertical peak strain magni-
tudes; it will be noted that there is nearly a one-to-one correspondence in
strain values, irrespective of the distance from the fracture path. Almost all
peak strain values fall into a range between O. 0008 in. /in. and O. 0030 in. #in.
The minimum principal strain traces are characterized by a shape that
-42-
MAXIMUM VERTICAL STRAIN — In, / In,
Fig. 88. Comparison of Maximum Prin-cipal and Vertical Strain Magnitudes.
exhibits a slight peak in tension,
followed by a fairly sharp com-
pressive pulse occurring as the
fracture propagates past the
rosette; this compressive pulse
corresponds to the tension peak
of the maximum print ipal strain.
The traces then return to the final
strain level of the rosette.
In the case of rosettes
located close to the fracture path,
the magnitude of the strain peak
is relativ~ly large. As the dis-
tance between a rosette and frac–
ture increases, the peak principal
strain magnitude decreases rapidly. This may be seen clearly in Fig. 20, in
which typical principal strain traces for rosettes located at various distances
from the fracture are presented. When the distance between the rosette and
the fracture exceeds about 2 in. , the rate of decrease of the peak strain magni-
tudes is somewhat less. Figure 89 is a plot of all the peak maximum principal
strains versus distance between the rosette and the fracture. This figure gives
the impress ion that there is a wide scatter; however, if the curves are plotted
for rosettes located at varying distances from the initiation edge ( 8 in. , 15 in. ,
etc. ), the individual curves move to the right with increasing distance, as may
be seen in Figs. 90 and 91. Thus, there is an increase in the peak maximum
principal strain with increasing distance from the initiation edge; this increase
is noticeable to a distance from 22 to 29 in. from the initiation edge, after
which there is no apparent change. Thus, the strain field associated with the
propagating fracture appears to reach a “ steady state” condition at about 22 to
29 in. from the initiation edge.
The variation of the maximum strain rate with vertical distance of the
..-
-.
—
.—
,.-
.—
—
w
DISTANCE BETWEEN ROSETTE AND FRAcTURE — in,
-44-
10
B
6
4
2
0
Ip.— ,’=f+
I
-r
i1“II\
\-,
;“:
\
.’\ .-I
.’\ “““ \ ,.
“. ---- \. ...-
,. .
0 10 20
‘ . .
-—~—. -— ___
-K’MAXIMUM STRAINRATE (6’)
—z TYPICAL <moq,
a
sw
~,TIME
.— __ .— _ _ —— ___
50 60 i
MAXIMUM STRAIN RATE (E’) – ..in, / in, /Sec.
FIG. 92 MAXIMUM STRAIN RATE VERSUS DISTANCE BETWEEN ROSETTE AND FRACTURE
rosette from the fracture surface was also studied. The strain rate (in./in./
see) is difficult to determine accurately because of the very nature of the
strain traces (Fig. 4); the computed strain rates presented in Fig. 92 should
be considered as only a qualitative picture. However, it is apparent that a
general trend of strain rate versus distance between rosette and fracture does
exist, as might be expected. The sketch in the upper right corner of Fig. 91
illustrates the procedure by which the strain rates were determined. A “best
fit” curve of all points is also shown On the plot. Strain rates ranging from 1
to 109 in. /in. /see were computed; however the majority of the values fall in-,.,.
the range of 1 to 30 in. /in. /see for gages located at 1/2 in. to 3 in. from the
fracture.
Figures 38--41 present strain-time curves of the component gages, as
well as the computed principal strains of the four rosettes for which records
—.
--
,-.
.—
. .
—.
-45-
~,ere ,secured, as taken from Test 35. This test was conducted at a higher stress
( 28, 000 psi) and a lower temperature (-15 F) to determine the effect of a higher
stire ss level on the s-train pattern. The traces are essentially the same shape as
those fou,nd in the plates te steal at lower stresses, although the peak maximum
s.tra ins. are slightly higher. However, the rosettes were located quite close to.-.
the fracture path and in view of the earlier
the stra, in magnitudes might be fairly high.---
magnitudes of rosettes obtained i.n Test 35
discussion it is to be expected that
The maximum principal peak strain
are plotted in Fig. 89. These exhibit
t“he same trend as peak strain magnitudes from the other tests. Thus, on the-..
basis of one test only, it appears that there was no marked change in the strain
belaavior for a propagating crack at this somewhat h}gher stress level..—
Maximum Principal and Vertical Strain Contours.—
In order to depict the strain distribution in a plate during the time a frac-
ture is propagating, it was decided to plot contours of certain strain components
for various crack lengths. The available data consist of strain–time traces for
horizontal, diagonal, and vertical gages obtained from this series of tests, as
well as from earlier tests, and computed. principal strain values for the rec-
tangular rosettes formed with the component gages of this test series. The maxim-
um principal strains. are of particular interest since they provid~ both the magni-
tude and direction of the maximum tensile strain.—
Strain contours were obtained @ superimposing the re suits from a number
of tests. The data were superimposed only from plates tested under similar con-—-.
ditions of stress, temperature, and impact; the only major variable among the
in,c!ividual tests was the location, of the fractu~-e path with respect to the, notch—.
line. In general, the fractures sloped upward from the point of initiation and
then leveled off as they traveled across the plate. The excursions from the notch
line were of the order of 2 to 8 in. ; precise locations of the fracture paths are
tabulated ~QFigs. 5--11. However, since these distances are small in relation
to the plate width, the actual crack length was never more than a few per cent
greater than the net plate width. The horizontal projection of the fracture was used
as the common fracture path On the ~trai~ contour plots.
.-
-46- .—
The procedure used in plotting the strain contours may be summarized as
follows . It was first necessary to estab] ish from the original strain-time curves
the time relative to various sel,ected crack lengths. W,so it was necessary to
establish the position of the gage with, respect to the fracture path. Then, at,
any particular time corresponding to a given crack length, the strain values were
obtained from the test data, and, these strain data were used in plotting the con-
tours . The contour lines were drawn in regions where data were available by
joining points of equal strain. No contours were drawn in regions where there
were no data, although it is obvious that strain, contours do exist in these re-
gions,
Before strain values could be determined for any particular crack I ocati,on,
it was necessary to establish the respective times that corresponded, to the desired
crack positions. in every test. The se times were determined by a knowledge of the —
breaking times of the crack detectors and the peaking times of the vertical strain
gages located closest to the fracture surface; they are tahdated in Table 5 forvari-
ous crack lengths. It was assumed that the peak strain of a rosette occurs when
the surface fracture is directly above or below the rosette, although the exact posi-
tion of the tip of the fracture cannot be determined at any precise time. At points
where there was disagreement between d,et.ector breaking times and gage peaking
times, the peaking times of the strain gages were selected as the more reliable in–
dication of the surface fracture location beca,u se of the more sensitive response of --
the strain gage. As a general rule, the detector breaking times were used only as
a guide in the selection of times corresponding to selected crack lengths. At frac-
ture positions where there were no strain gages or detectors, the corresponding
time was approximated by a linear interpolatim between any two known points.
Since the horizontal projection orI the notch line of the actual fracture path
was used as the common fracture path in all tests for purposes of superposition, it
was important to locate accurately each gage or rosette with respect to this com-
mon fracture. Three different methods of d.etermininq the gage and rosette locations ‘-
on the plate layout using the common fracture path were investigated. The first
method consisted of measuring the vertical distance between the rosette and the
actual fracture for particular tests and locating the rosettes on the plate layout ac -
-. . -47-
TABLE 5. TIMES CORRESPONDING TO VARIOUS CRACK LENGTHS,.—
—-
-..
.—I
—.
—
All times. are in milliseconds and refer to the time base of thestrain-time curves presented in Figs. 21--85
CrackLength(in. )
Test No.
33 34 37 38 39
8
15
18.5
22
25.5
29
32.5
36
39.5
43
0.22
0.94
1.05
1.16
1.24
1.32
1.57
1.62
1.74
1.86
0.03
0.50
0.7’2
0.95
1.10
1.25
1.37
1.50
1.63
1.77
0.92
1.11
1.18
1.24
1.35
1.4”5
1.56
1.66
1.82
1.97
0.35
0.52
0.61
0.70
0.81
0.92
1.03
1.14
1.26
1.37
0.10
0.31
0.45
0.59
0.67
0.75
0.91
1.07
1.18
1.30
Note: These times, corresponding to various crack lengths, were used in determin-ing both the maximum print ipal strain contours (Figs. 93-- 102) and the verti-cal strain contours (Figs. 1.03–– 112) .
cordingly; the second method consisted of measuring the perpendicular distance of
each rosette or gage from a horizontal line projected across the plate from the tip of
the surface fracture for the particular crack length under investigation; and the third
method consisted of 1) measuring either the perpendicular distance between a rosette
and the fracture or the extension of the fracture in the direction in which it was travel-
ing at that particular instant,
fracture path by this amount.
rosettes on the strain contour
and. 2) locating the rosette above or below the common
A study of the three methods of locating the gages and
plots. indicated that regardless of the method used, the
final strain contours were essentially unchanged. Since, in general, the fracture fol-
lowed a horizontal path as it passed through the region where rosettes were placed, it
was obvious that in the majority of cases there would be no significant difference in
-48- ..=
the final rosette locations as determined by any of the three methods investigated.
In tests where the fracture was not horizontal, there would be a change in rosette
locations; however, the changes in strain at some distance in front of the tip of the
fracture were usually so small that the strain contours were still not altered markedly.
Therefore it was decided to use the actual fracture path as the common reference
line for location of the strain gages and rosettes.
The maximum principal strain contours are presented in Figs. 93--102, and
are based on the data from Tests 33, 34, 37, 38, and 39. The data from Tests 35
and 36 were not used in the contour plots because Test 35 was conducted on an ar–
rester specimen at a higher applied stress level and Test 36 resulted in a double
fracture. However, careful examination of the strain-time records (Figs. 38--52)
from gages and rosettes of Tests 35 and 36 will show that the data from these two
tests are similar to those of the other tests reported. In the contour plots, t“he mag-
nitudes and directions of the maximum principal strain are shown at the ~espective
locations on the plate layout; the directions of the maximum principal strain are
shown as short straight 1ines. For ease in interpretation of the contours, all strain6units are in. /in. times 10 . Since all of the tests were made at about the same
stress level, it appeared desirable to plot the contours in terms of absolute strain,
i. e., with respect to the as-rolled condition.
In the majority of cases, as the fracture approaches the rosette, the direc-
tion of the maximum principal strain rotates slightly, so that it points toward the
approaching fracture, and then returns to a more or less vertical position. as the
peak value of the strain occurs. After the maximum strain value is reached, the
direction of the maximum print ipal strain continues to rotate and, in general, con-
tinues to point toward the surface fracture tip. Values of @, plotted. as short
straight lines in the direction of ~ on the principal strain contours, show thismax
behavior quite clearly.
For an 8-in. crack (Fig. 93), the shortest crack length for which contours
have been plotted, it will be noted that the increase in strain directly above and
below the surface fracture tip is small in comparison with strain changes for longer
crack lengths. At points in front of the 8-in. crack, the strain changes are negligi-
ble and the plotted strains correspond to the test load principal strains. The varia -
.. .
—
.-.
-49-. .
-. .
—
–8
-16
+ 000133-clNOTE, TEST LOAD $TRAIN w +600.
$TRAIN UNITS – lN./l N. X 10$
k
TfST HO.POSETTE MO
b 660 (37-71+ 470t3a-r) I 570 ,sm-#)
+ 730 (37–8)+ 670(,,–6)
/
630 (34-1)*OO ‘ 760[s”-cl \ 620 (37-51 + 560{30 -41 I 560 [30-.1
‘“00?-737-4’
660 (37- z! ~ 640133-m
J
.,0(37-, j ::;~+ 1 i%izl
t 570,,, -., 4 560130–21 , 4eo (s,-!)
——- ——..—.62013+3)
— —~ :gg E,::y b 600 (3s–.1 i +To (30–3,
#B50,,,-,. + ‘50(37-” , 6,.,,,.’,, t s40(,+c~ / :::::::; * 370,3*-10) , ,c.0(,,.,,
b% , B20(,4. AW), ~,~,$,.z) 5001.9–9), ~90,,,.9) , 590,3.-8) ; 5601=–31)
5@o(34.6n K1
+ 590(39-ml i 600[39–0>
~ w: 70 “= NET PLATE WI.,.
0 0! a2 03 04 0,5 06 0.7 0,0 a9 1.
FRACTION OF MET PL4TE WIOTH (W)
. . . . .- .= AX .,l ●7 CA
DISTANCE FROM INITIATION EDGE — IN,
FIG. 93 MAXIMUM PRINCIPAL STRAIN CONTOURS FOR 8 — IN. CRACK
,1+ 030
NOTE. TEST LOAD sTRAIN w +600.
STRAIN UNITS – lN/l N. x 106
I
I 500
+ 560
4 460
I 560
~ w“IO”= NET PLATE wIDTH — --q
1 ~, -~o0 01 02 03 0.4 0,5 06
FRACTION OF NET PLATE WIDTH (W )
8 15 22 29 36 43 50 57 64 72
DISTANCE FROM INITIATIoN EDGE — IN,
FIG, 94 MAXIMUM PRINCIPAL STRAIN CONTOURS FOR 15 — IN. CRACK
-50-
tion in test load, print ipal strains for the rosettes and for t“he various tests accounts
for the differences noted in the plotted strain values. The numbers in parenthesess
following the strain values refer to the test number and come spending rosette from
which the plotted. strain values were obtained, These t-wo numbers are shown only
on Fig. 93 but refer also to Figs. 94--1.02 as well.
For an 8-in. crack. length, the small changes. in strain exhibited by the
rosettes located only 8 in. from the initiation edge attest to the fact that the notch-
wedge- impact method of fracture ini.t iat. ion has little effect on the strain distribution
in a wide plate at some distance from the source of initiation.
The change in strain distribution as the crack length increases may be seen
clearly by comparing Figs. 93, 94, and, 95, in which the maximum principal strain
contours are presented for crack lengths of 8 in. , 15 in. , and 18.5 in. As the crack
length increases it is noted that the rnagnitud,e and extent of the strain field, associated . .
with the crack tip also increase. At a crack length of 22 in. (Fig. 96), the strain pat-
tern surrounding the moving crack tip ceases to change, and for crack lengths in ex-
cess of 22 in. (Figs. 97-- 102), the strain field surrounding the advancing crack tip
remains essentially unchanged. This general effect was noted earlier in connection
with Figs .90 and 91, which show the results of a study of the peak strain magnitude
versus distance of the rosette from the fracture for rosettes located at various dis- .
tances from the initiation edge. The extent of the strain contours directly in front of
the fracture increases only slightly with increasing crack lsngth. Thus, for this
particular specimen geometry, the propagating brittle fracture apparently does not
reach a “ steady state” condition until it, has traversed a distance of about 22 in. -.
This length is somewhat longer than that sugge steal by earlier studies.
As may be seen in Fig. 96, the major portion of the strain field associated
with the propagating brittle fracture extends only about 8--10 in. in front of the
fracture. In the region directly in frontof the cracks, the strains decrease slightly
as the fracture approaches. It is likely that some of the strain variation observed
directly in front of the fracture is associated with the formation of fracture nuclei
in front of the main fracture, or with transmitted strain changes re suiting from the
fracture and strain energy relaxation.
The main changes in strain occur above and below the fracture tip and
—
—
.
-51-
1I—————v--——————— -.
I
~, .
.—— ————— .. ———.—. ————————-. . +
4? i
,—01
‘Ml - 3UflL3VUd M0130 UO 2A00V 3W4V1$I0
—
...
-.
..-
-. ——.. ....—g .
—.———.—-g .
UO 3AOEV XINVLS1O,NI — 3Bn13VW3 M013e UO 3A0QV 33 NV1SI0
...
FIG 99 MAXIMU!& PRINCIPAL STRAIN CONTOURS F03 32.5 — IN CRACK
I ,50m,, ,.,, ,... ,,.,,” . ..s.0
s,”, “ “,1,s — ItL/!* x m’
“ .,0,,. “ET?,,,, .,.,.
I0 0 )5 22 29 36 43 59 57 64
DISThtlCE FMOM 1?4lTL4T10N EDGE — IN,
FIG, !01 MAXIMUM P. WNCIPAL STRIMN CONTOURS FOR 395 — IN. CRACK
Ir
I
,0 e 15 22 29 36 43 50 57 64
01 STAf4CE feofd IUITIATIOf4 EDGE — w.
FIG. 100 MAXIMUM PRINCIPW STRAIN CONTOURS FOR 36 — IN, CRACK
Ar-d
I
–e
–16
, 30. /, ,W
.- 10-- .[1 PLATE WI*T”II
I. 0, .2 0, FWrlo,”m “,, %::, W,n,” ::,07 0, 0,
0 e. 15 22 29 36 43 m 57 =4DISTANCE F40M dN4r1hT10N EDGE — IN,
FIG 102 MAxIMuM PRINCIPAL STRAIN CONTOURS FOR 43 — IN. CRACK
I I
-53-
sligbtly in front of it. Behind the fracture front the strains decrease rapidly. Ac -
tually all strain contours appear to converge at the fracture front. Although ex-
tensive strain measurements were not obtained at distances far above and below
the fracture, the recorded strain data indicate that the contours are symmetrical
around t“he fracture path.
Further information on the extent of the strain field. associated with a propa-
gating brittle fracture may be obtained by studying the strain traces presented, in
Figs. 44 and 46, These strain traces are for two rosettes (Numbers 3 and 5) lo-
cated between the double fracture occurring in Test 36 {Fig. 8) . Each rosette v7as
located approximately 2 in. from one of the fractures and the strain traces of the
rosettes are typical of strain curves for rosettes so located. A study of the strain
is
data would seem to indicate that the strain field associated w-ith either of the ad-
vancing fractures was affected little, if any, by the strain field associated with
the other advancing fracture.
As has been notdd earlier, the magnitude of the maximum print ipal strain
essentially equal to that of the maximum vertical strain; one reason for this is that
the principal strain direction does not rotate much from the vertical. Thus, as a
matter of comparison and because a plot of the vertical strain a,llows an incorpora–
tion of data from the earlier tests, it was decided to pre sent cQntours far the vertical
strains as well; these are presented as Figs. 103--11,2. It is to be noted that in
these figures the reference strain is the test load strain.
The vertical strain contours could have been piotted in terms of absolute
strain as was done for the print ipai strain contours. However, in order to incorpo-
rate the data from earlier tests in which the test load stress varied from 15, 000 psi
to 20, 000 psi, it was decided to subtract the test ioad strain from. the absolute verti-
cal strain and plot the relative strain values. In Fig. 103 the numbers in parentheses
following the strain values refer to the original test number and corre spending vertical
gage number respectively from which the plotted strain values were obtained. In gen-
eral, the. vertical strain contours are similar in s“hape and magnitude {if the test load
strain were added) to the maximum principal strain contours discussed previously.
With regard to the strain data, which are plotted at any particular time for
purposes of establishing the strain contours, there are severai additional factors of
I
Ic
e
?OO
T-)($7-101 2 .
‘3’-’5] woo
0. 133.7)NOTE: REFERENCE STRAIN IS TEST LOAD STRAIN
STRAIN UNITS - lN,/l N X 106
/’+TEST No.
VCFf TICA L C40E MO
0. (37.20)I
+o.<~.-z.l C. 138-3(1
0. (37.3,)60. {37-5)
0. 13..13) 0. (3. .,,).s~.!37-23) 0.(3.-,01
0.(37.8) 0.(33.4)0.<39-281 0 (39~31)
0.(3? -31
0.(39-5)O: 134-101 ;:\::;; 0.(3a-010 (33.10)—
o.(33-la 25) ~,<3,.,o, 0. (99<s70.(37-13) .Go.\9T-lol
0.(3.-71 .40. (3.-13)0.(3, -s1 , ,0:197-20) 0.1%0-13)
~ (33-2UI0.1,*-,.,2B ~.<3m..l 0. (39-131~, [39-,01 0.(34.41 .60: 138-201
0 13-.16931)
c. 1s9 -51
c.{3e -3)
~. 138-101
,3.1$,-,51 0. [39-201
r W = 70””~ NET PLATE WIDTH ~
I I
o 01 0,2 03 04 0.5 0,6 0,7 O.a 0.9 1.FRACTION OF NET PLATE WIOTH (W )
o 8 15 22 29 36 43 50 57 64 7
DISTANCE FROM INITIATION EDGE — IN.
FIG. 103 VERTICAL STRAIN CONTOURS FOR 8 –IN. CRACK
16
8
-a
-1 f
140.NoTE. REFERENCE ST R41N IS T:ST LOAO 57 RA1N
STRAIN uN(?S – Iwlv x 106
120-
~~vhd -’!;-30. 0.————-
/
50.-50 . .10. -30 . –
020--600. -0.
\
0, -7040 :-50
0,~@
0.0. -6; ;
\o 0. eo .
0-
’30.
-40.
-60.
0.
0,
—0
0.
h w’70” ‘ET ‘LATE WIDTH ~
I I
o 0.1 0,2 0.3 04 0.5 0.6 07 00 09 1FRACTION OF NET PLATE WIDTH (W )
o 8 15 22 29 36 43 50 57 64 ;
DISTANCE FROM INITIATION EDGE — IN,
FIG. 104 VERTICAL STRAIN GONTOURS FOR 15 – IN,CRACK
-55-
———————————.— —
I Tl”l
2
1.
.-
‘------’” ‘TI
. .
g7 1MO 3AOEV 33 NvLS,0
-56-
--------
————.— .—— ——-——. . I t’
“1
!————— ———————.
-57-
irnportance that should be noted. Because of the nature of the recording equip-
ment, the component gage strain and the computed principal strains are accu-
rate only to about the nearest O. 0001, in. /in. , and the recorded times are accu-
rate only to about O. 05 milliseconds. For an average fracture speed of 2500 fps,
this time error of O, 05 milliseconds would. correspond. to a fracture length of
about 1– 1/2 in. The value of the test load strains varied somewhat for the tests
because of the slight bending of the plates which mms caused by the welding of
the specimens to the pull-plates, and by changes of the residual strains in the
plates . Consideration of such variables led to the decision to not plot contours
at intervals of less than O. 0002 to 0.0004 i,n. /i.n.
A set of typical maximum principal strain contours is presented in. Fig.
113; these contours are based on the results presented in Figs. 96--1o2, and,
for the geometry and test conditions reported here, are considered to
sentative of the strain field associated with the crack tip for a crack
excess of 22 in.
be repre-
length in
A typical set of vertical strain contours based on the results of Figs.
106–-112 and for a crack length in excess of 22 in. , is presented as Fig. 114.
Vertical strai~ contours for a crack length of 29 in. , based on results
from all earlier 6-ft wide plate tests are presented in Fig. 115. A tabulation of
the data used to plot the contours in Fig. 115 is, presented in Table 6. These
values are typical for the crack lengths in excess of 22 in. investigated for
Tests 13--32. The strain va~ues ‘were superimposed in a manner described
earlier; although a few exceptions may be noted, the strain values exhibit the
same general pattern as shown in Fig. 114.
Although contours of only the maximum pri~c ipal strain and vertical
strains are presented in this report, contours of the other strain compo~e~ts,
namely the diagomi strain, lmrizo~tal strain., mi~imum principal strain, and
shear strain, could be prepared from the strain data presented in Figs. 21--85.
The contours of maximum print ipa,l strain .ard vertical strain are based
on data obtained from all the 6-ft wide plates t~sted as a part of this investiga–
tion, and are considered to be representative of the strain distribution recorded
on the surface of the plates dm-ing britile fracture propagation. In spite of the
TABLE 60 DATA OBTAINED FROM PREVICIUS TESTS 13-32(See Fig. 115)
VerticaI Distance Distance Time Test StrainTest Gsge of Gage from of Gage from Comes. Total Load Referenced toNc?a No. Init. Edge Fracture to 29-in. Strain Strain Test Load
Crack Strain(in. ) (in. ) (milliseconds] {in. in. ) (in. /ino) (in. in. )
13 1
23
15 I23
19 24
22 14
23 56789
25 12349
32 1234
12.0
35.5600035.0
36.036.036.5
61.037.0
61,036.043.050.057.0
64.08.0
1!5.0
22ao29.0
64.015.036.0
50.057.0
-0.3
-1,7
-2,01,61.6
-3.4
-0.60
-3.5
-3.5-3,0-3.4-3.5-3.5-3.5
==0.5-0.8-0.7
-0.7-0.8
-2.0-6.6
-7.2-7.1
0.66
0.660066
0.810.810.81
0.560.56
0.320.32
0.820.820.820.820.820.920.920.920.920.92006000600.600.60
.00063
.00072
.00066
.00079
.00071
.00090
.00052
.00055
.00082
.00062
.00067
.00067
.00064
.00063
.00052–.00003-.00001
-.00007.00231.00061
-.OOO11.00095
.00050
.00054
.00063
.00072
.00066
.00064000049.00078
.00047
.00055
.00064
.00062
.00067
.00067
.00064
.00063
.00062
.00059000061000061.00056
000061.00057
.00055
.00055
.00054
0
00
.00015
.00022
.00012
.00005
0.00018
0
00
00
-.00010-.00062-000062-.00068
.001750
-.00068
.00040-.00005
0
24
–16
FIG. 113
NOTE. STR41N UNITS — i., /in, X 10s.
TEST LOAD STR41N = 600
,~/f!-@?rj~6001000
600 800Eoo
6 m
–12 –e -4 0 4 a 12 16 20 24 28 32
DISTANCE AHEAD OR BEHIND OF SURFACE FRACTURE TIP — in.
TYPICAL SET OF PRINCIPAL STRAIN CONTOURS FOR 22-in. TO 50-in. CRACK
24
2C
IE
12
8
4
c
-4
-a
–12
-16
FIG. 114
NOTE , REFERENCE gRAIN IS TEST LOAD STRAIN,STRAIN UNITS — ;“./i.. X 10’
.../-j/V>\0
200 0
n m-12 -s –4 04 B 12 16 20 24 26 32
DISTANCE AHEAO OR BEHIND OF SURFACE FRAcTuRE TIP .— in.
TYPICAL SET O F VERTICAL STRAIN CONTOURS FOR 22 – in. TO 50- in. CRACK
—
-60-
diiOTE: REFERENCE STRAIN IS TEST LOAO STRAIN
STRAIN UNITS - lN.flf> X 10S
s16
uu Iza+go
“6 “‘~ w =70”=NET PLATE WIDTH
0 0,1 0,2 0.3 0,4 05 0.6 0.7 0.8 0.9 Id
FRACTION OF NET PLATE WIDTH (W 1
0 s 15 22 29 36 43 50 57 64
DISTANCE FROM INITIATION EDGE — IN.
FI13, IIS vERTICAL STRAIN CONTOURS FOR 29 - IN. CRACK -— TESTS 13-32
limitations discussed in this section, it is felt that ths strain contours pre-
sented give a very representative picture of the effec-i of a propagating brittle
crack on the strain distribution in a wide steel plate.
Crack Path and Surface Texture——
Photographs of the fractured plate specimens a:-e presented m Fig. 116.
In general, the fractures slope upward from the point of initiation and then level
off. A black string is stretched along the notch line ir~these figures. IrI most
tests a secondary crack started from the notch cut at the far edge, propagated
toward the approaching fracture, and ended in a subme;ged crack before reach–
ing the main fracture. These secondary cracks may be seen in the figures. The
fracture in Test 36 branched into two complete fractures near the center of the
plate.
Views of a typical fracture surface are presented in Fig. 117. The shear
lip associated with the fractures was very small and almost imperceptible. Por-
.—
-61-
Test 33
— ,,, ,— ,,, — “-”:-““”w
,-:<, .--: ...’!“’!.)<1,\;,
‘~!’‘!...fl !“’
Test 35
Test 36
,,>,, ,. , !,,,, >,L”L,Z:’ “ ‘“!’ “,~ik>,.,”,..,,.,.. . .LL!L.L.L, .-. .-- . .. —.-
Test 37
Test 38
, ,,.,. ,,.,,,,,,i ‘,’”;ie-——-.—. .-. .“.4+7 ..-..7A.,
d
—–--~J -----
-..--”.
Test 39
Fig. 116. l%acture Paths--Tests 33--39
tions of the surface exhibit a coarse texture with the usual herringbone pattern,
while in other regions, usually ~.ear the initiation edge, the texture is quite
smooth. Enlarged views of typical smooth and coarse fracture texture are pre-
sented in Fig. 118.
The reduction in plate thickness in the region of the fracture surface
was one to two per cent; this reduction in thickness is of the same order of
magnitude as that found in earlier fracture tests conducted as a part of this
program.
-62-
~A, ., .7,- >.——
Fig. 117. TYPical Fracture Surface--7est 3S
Fig. 118. View of Portions of Typical Fracture Surfaces fromTest 34 Showing a Coarse Surface on the Left and a SmoothSurface on the Right.
SUMMARY
This research report contains a summary and preliminary analysis of
data primarily from Tests 33-- 39, conducted as a part of the current brittle
fracture propagation study. More complete studies of these data, as well as
data obtained subsequently, are being made as a part of the program.
The specimens were tested with a net stress o! about 19, 000 psi, at a
temperature of about -10 F and with a rmm.inal impact af 1200 ft-lb for fracture
initiation. Thirty-four channels of instrumentation were used for each test;
this is more than three times as many channels as hai previously been availa-
Me for any one test. The strain-time traces from the component gages, as
well as the computed principal strains, are presented. Since the tests were
-63-
--
conducted under similar conditions of stress, temperature, and impact, it was
POSsible to superimpose the strain data to obtain a more complete picture of
the strain distribution on the surface of the plate associated with the moving
fracture.
The behavior exhibited by the component gages was similar to that ob-
served in earlier tests. For example, in the case of a vertically oriented strain
gage, as the distance between the gage and the fracture path decreased, the
strain pulse became sharper and had a greater magnitude; as t-he distance be-
tween the gage and fracture increased, the strain pulse extended over a longer
period of time, but the precise shape of the pulse depended on the distance
from the fracture path. For gages oriented in the diagonal and horizontal direc-
tions, this same general trend was noted, but the shape of the strain trace, par-
ticularly in the case of the horizontal gage, was markedly different.
In the case of the principal strains computed from the component strain
traces, the maximum principal strain was found to exhibit the same type of be-
havior as observed for a vertically oriented gage, with the exception that the
direction of the maximum principal strain was found to rotate slightly to either
side of the vertical axis depending upon the location of the fracture.
Although the nature of the strain traces are such that only a rough pic-
ture of the strain rates may be obtained, the trend is of some interest. As
would be expected, the maximum strain rate occurs for gages located quite
c~ose to the fracture. Strain rates ranging from 1 to 109 in. /in.. /see have been
computed; the majority of the values fall in the range 1--30 in. /in. /see for
gages located 1/2-- 3 in. from the fracture.
Contours of maximum principal strain and vertical strain for various crack
lengths are presented. The major portion of the strain field associated with the
front of the propagating brittle fracture extends only about 8--10 in. in front of
the fracture; the regions most affected are above arid below the fracture tip and
slightly in front of it. Although extensive strain measurements were not obtained
at distances far above and below the fracture, the recorded. strain data indicate
that the contours are symmetrical around the fracture path.
A study of the data reveals that for crack lengths in excess of about
-64-—
22 in. , the extent, magnitude, and nature of the strain field associated with the
advancing tip of the fracture remains essentially uncharged. Typical maximum
principal strain contours and vertical strain contours based on all of the availa-
ble data are presented for fractures propagating in the central one-third of the
plate width.
REFERENCES
1. Hall, W. J., Godden, W. G., and Fettahlioglu, 0, A. , Preliminary StudiesQf Brittle Fracture Propagation g Structural Steel (Ship Structure CommitteeReport Serial No. SSC- 11 1), Washington; National Academy of Sciences-National Research Council, May 15, 1958.
2. Lazar, R., and Hall, W. J., Studies Q .Brittle Fracture Propagation in Six-Foot. Wide Structural Steel Plates (Ship Structure Committee F?eport Serial
.—
No. SSC-1 12), Washington: National Academy of Sciences-National I?e-
search Council, September 17’, 1959,
30 Lynam, T. M. , Strain Patterns During the Propagation of a Brittle Fracturein Steel Plates (M. S. Thesis submitted to the Graduate College, University——of Illinois), 1957.
4. Rolfe, S. T., Studies of the Strain Distribution in Wide Plates During Brit-——— ——tle Fracture Propagation (M. S. Thesis submitted tc the Graduate College,~iversity of Illinois), 1958.
—
—.
-.
5. Hall, W. J., Mosborg, R. J., and McDonald, V. J ~, “Brittle Fracture Propa-gation in Wide Steel Plates, “ The Welding Journal, 36:1, Research Supple-ment, pp. 1s-8s (January 1957~
—+
Chairman:
COMMITTEE ON SHIP STRUCTURAL DESIGNDivision of Engineering & Industrial Research
National Academy of Sciences-National Research Council
;’.
,..-.~I]:
,.-i
I-’
I ‘-
N. J. HoffHead of Department of Aeronautical EngineeringStanford University
Vice Chairman:——
M. G. ForrestVice President-Naval ArchitectureGibbs and Cox, IrIc.
Members:
George F. CarrierGordon MacKay Professor of
Mechanical EngineeringHarvard University
C. O. DohrenwendDean of Graduate SchoolRensselaer Polytechnic Institute
J. M. FranklandConsultant, Mechanics Divi$ionNational Bureau of Standards
Dana YoungDean, School of EngineeringYale University