LABORATORIUM VOOR
SCHEEPSCONSTRUCTI ES
TECHNISCHE HOGESCHOOL - DELFT
RAPPORT Nr.
SSL 171
BETREFFENDE:
Crack initiation- and propagation-tests on welded joints of st. 52.
SHIP STRUCTURES LABORATORY Deift University ¿f Technology. Report no.
SSL 171
4.December 1972.
CRACK INITIATION- AND PROPAGATION-TESTS ON WELDED JOINTS OF ST. 52
by
CONTENTS:
page
Introduction 3
Data about plate metal and weld 3
Main.test-specimens and test set-up '4
'4. Additional tests '4
5. .Results of the experiments.
5.1. Static C.O.D. bending tests '4
5.2. Fatigue bending 4
53.
Observations in connection to the resultsof the tensile tests and the Charpy-tests 5
Conclusions 5 Literature 5 Appendix I 7 Appendix II 9 Appendix III 10
-2-I
CRACK INITIATION- AND. PROPAGATION-TESTS ON WELDED JOINTS OF ST. 52. By J.J.W. Nibbering and R. Vonk.
1. Introduction.
One problem in connection to welded butt-joints is whether a large difference
in yield strength between plate- and weld-metal is acceptable or not.
Another problem connected to this, is whether a too high ratio between the
yield point and the tensile strength of the plate metal (and/or the weld metal) has to be avoided as is suggested by certain codes.
The point of view of the authors is that in principle both aspects are not
of significance in relation to brittle fracture, provided that the
notch-toughness of the materials at sub-zero temperatures is sufficient /1/.
This property is mostly controlled by. means of Charpy-V-notch impact testing.
For normal cases this test may be sufficiently reliable but for special cases,
like the one considered here (high yield points of parent plate- and/or weld metal) better tests are desirable.
For statically: loaded conditions C.O.D.-testing (C.O.D. crack-opening
dis-placement) has become an excellent method for estimating the resistance of
metals to crack-initiation at the tip of a sharp notch or defect. When this
resistance is sufficient, in principle no need exists for determining also
the resistance of such metals against high-speed (brittle) crack-propagation.
However with the aid of crack initiation tests 100% safety against initiation
can never be obtained. This is connected to the fact that at defect-tips in
structures local embrittlement of the metal can occurr in a
difficultly-to-reproduce way /2/. Consequently there is always a very small but real chance
that a brittle crack initiates in a small embrittled region. Then the adjacent
material should be able to arrest that small crack. A suitable test for the
estimation of the service temperature above which that will occurr is Pellini's
drop weight test. Unfortunately it is not very suitable for welds and
heat-affected zones (H.A.Z.). Therefore as a substitute, the authors propose
low-cycle fatigue loading at low temperature.
In these circümstances tiny brittle steps may develop in the locally damaged
material at the tip of the fatigue-cracks. When these indeed develop one or
more times, the material proves itself to be able to arrest these tiny brittle cracks.
Both types of experiments: static C.O.D. and fatigue bending at low
temperat-ures have been carried out with specimens of 12 and 28 mm thickness. The cnt-ical (minimum required) C.O.D. was 0,3 mm at -20°C for weld and H.A.Z.
2. Data about plate metal and weld. .
All information about the welding (sequence etc.) and the properties of the weld and the plate metal can be found in appendices I, II and III.
L
'ç
Main test-specimens and test set-up.
The specimens are shown in figure 1. The notches for the static C.O.D.-spec-imens are of the standard form used by the Ship Structures Laboratory.
The notches in the fatigue-bend specimens were of smaller length than those for the static bend,specimens. In this way a long path for the development of fatigue-cracks became available. The position of the notches is indicated in the figures of table I and II.
All specimens were loaded in four-point bending. For the statically-loaded C.O.D.-specimens the load was increased with steps of 2 tons after which the C.O.D. was measured. For the 28 ¡mn specimens the test temperature was -20°C in case the notches were situated in the weld or H.A.Z. and -30 C when the notches wege in the parent plate. For the 12 mm specimens the temperature was always -30 C.
The specimens loaded in fatigue bending were regularly observed when the fatigue cracks propagated. When a complete fracture had not yet develòped at the moment one third of the height of the specimen had cracked, it was loaded to fracture, in order to be able to inspect the surface for the presence of eventual brittle steps.
Additional tests.
In order to know accurately the yield points and tensile strengths of the investigated welds and Heat-affected zones, small tensile bars have been
taken from the material and tested. (See fig. ti.).
Furthermore Charpy-V-notch specimens have been fabricated and tested,.(fig. ti.a).
Results of the experiments.
5.1. Static C.0.D. bending tests.
11+ Specimens have been tested; 6 with a thickness of 12 imn and 8 of 28 mm. Figures 2, 3 and table I summarize the results.
o o
All specimens have met the requirement of 0,3 mm C.O.D. at -20 C (-30 C for the parent plate material).
Only the results of specimens RTD-21 and RTD-29 were relatively low, although greater than 0,3 mm. This may be connected to the fact that these specimens have been tested relatively soon after welding as compared to the other ones (3-4 days).
5.2. Fatigue bending.
5
with 28 mm thickness.
In table 2 the results have been summarized.
In all specimens, brittle steps did not even develop, which points to a very
high resistance against crack initiation. One might conclude that due to that no information is obtained with regard to the resistance of the materials against propagation of cracks. But a more reasonable and realistic conclu-sion can be drawn when the large experience with this kind of testing in the Ship Structures Laboratory is taken into account. This conclusion is that when no brittle steps occur under the circumstances the toughness of the material is certainly enough for arresting eventual tiny brittle cracks. This conclusion has been substantiated by the results of the final tests to fracture of the fatigue-cracked specimens. In all cases large "shear lips" developed along the edges of the surface during the final static loading to fracture.
5.3. Observations in connection to the results of the tensile tests
and the Charpy-tests.
In Appendix I the ratios between yield point and tensile strength of the various metals have been given. In most cases they are lower than 0,8. Only the weld fnetal for the 28 mm plates shows a value of 0,842. Moreover the value of 0,78 for the 12 mm plate material is also rather high.
From the foregoing paragraphs it will have become evident that these high yield-tensile ratios are not objectionable at all for a crack-safe
behavi-our in practice. This also applies for those cases where the difference be-tween the yield point of the weld is larger than 1,29 times the yield point of the plate-material connected. For the 28 mm weldments it amounted to 53,1
- 1 47
36,1 '
The results of the Charpy-tests at -20°C are relatively low for weld centre and H.A.Z. But in combination with the values at O C and of course the results
of the C.O.D.- and fatigue-bend-tests they do not give rise to special consid-eration.
Conclusions.
The.results of sophisticated experiments prove .that the high ratios between yield point and tensile strength of the metals tested or between the yield points of the tested welds and connected plates is not objectionable at all from the viewpoint of safety with regard to brittle fracture.
Literature.
/1/ J.J.W. Nibbering:
"Plastic deformations at notches in welds of mild steel plates". Report no. 129 S.S.L. Delft, Jan. 1968; 11W-doc. 2912-107-68/X--635-71.
Idem - Part II: "Notches in transverse welds"
Report no. 129a S.S.L. Delft, April 1970; 11W-doc. X-636-71.
s..
.
-6-/2/ J.J.W. Nibbering:
"Fracture mechanics and fracture control for ships". In: Lecturers notes of the K.I.v.I. course on
Engineering Fracture Mechanics, Eindhoven,
16-19
May 1972.-7
APPENDIX I.
Properties of plate material. (From Certificat Bureau Ventas).
For 28 mm thickness;. Chemical composition %: Cr Cu Mo Nb Al Cequ. Cu + Ni + Cr + Mo 0,016 0,022 < 0,010 . 0,032 0,070 0,L116 0,071 Mechanical properties: Roll direction transverse For 12 mm thickness; Chemical composition %: C Si Mn P S Al Nb 0,18 0,31 1,3t 0,019 0,01' 0,050 0,033 Mechanical properties: Roll direction
Yield point Tensile strength Elongation
(a kg/mm2). (kg/mm2) dp5 % 38,3 53,3 31,3 Yield point (a kg/mm2) Tensile strength (kg/mm2) Elongation dp5 % '45,0 57,8 29,1
-From attest bureau "Ventas"; 28 mm thickness (plate material):
a
a 38,3 kg/mm2 afract 0,718.
a fr
Tensile tests from Ship Structures Laborator 28 mm thickness (plate material):
.
-8-a I' 36,1 kg/mm2 afract 52,75 kg/mm2 O,68'. a fr Weld metal: o 53,1 kg/mm2fract. 63,2 kg/mm2 i--a O,8'2.
fr
From attest bureau "Ventas"; 12 mm thickness (plate material)':
= '5 kg/mm2 57,8 kg/mm2 ...L.. 0,78
5 SEQUENCE WELDING PARAMETERS ELECTRODE: 024830 Lot :7278 DIAM. ELECTRODE: 3,17
-9-APPENDIX: b'o
o
I-300 500 Pass N° Intensity A VoLtage V Energy kg/cm 1 120/130 23/24 25/30 1 II 3 IIIl
Il -4 1101120 5 120/130 ,,SEQUENCE
WELDING PARAMETERS
WIRE: FASCO 79S f12110
GAZ: ATAL FLOW: 24Lfmn /29L/mn.
CONTROLE: RX 100% : GOOD
lo
-APPENDIX: fi
non consumabLe strip
VERTICAL AUTOMATIC MAG PROCESS : VERTOWELD
260mm sb° BEVEL 28 Intensity A VoLtage V Travel speed mm/mn Energy kg/cm Preheat °C Pass N! 165/175 18/19 60 33 100 4 h 60 33 3
'I
60 33i'
2 62 32i'
I
160/170 19 - 90 18/20'i
5 18 70 23 6 18 65 25 7I
E Ee
g
H'
0.
TABLE - i (STATIC BEND TEST)
WII xoc
4o-froJur
,
o avSpecimen No. Thickn. (mm) Position of Notch C.0.D. (mm) Test Temp. ( C) Remarks RTD-1 12 weld 0,800 -26 -2 12 weld/fusionhine 1,120 -29 -3 12 average fusionhine 0,910 -30 -tj. 12 fusionhine/plate mat. 1,110 -31 -5 12 plate material 1,000 -31 -6 12 plate material 1,210 -28
RTD-21 28 weld )O,260* -20 tested within
-29 28 weld >0,280e -18 L8 hours after
welding
-'O 28 weld 0,580 -20,5 tested after
28 weld 0,380 -22 3- days after
-'.12 28 weld 0,700 -21,5 welding
-23 28 weld/fusionline 0,730 -19
-25 28 average fusionline 0,870 -22,5
s.
TABLE - 2 (FATIGUE BEND TEST)
Specimen No. Thick-ness (mm) Position of Notch Test Temp. ( C) Load (tons) Number of Cycles to Crack Initiation Number of Cycles for 8 mm Crack Length Brittle Steps
RTD-ll 12 weld -20 3-8 11.200 I8.5OO none
-13 12 weld/fusioni. -20 3-8 5.100 '39.900 none
-15 12 average
fusionl.
-20 3-8 5.000 38.000 none
-17 12 plate mat. -30 3-8 LI.600 26.100 none
N for 19 mm Crack Length RTD-30 28 weld -20 3-19 2.900 17.100 none -37 28 weld -20 3-21 6.000 -31 28 weld/fusioni. -20 3-19 3.000 32.000 none -33 28 average fusioni. -20 3-19 1.600 26.'3OO none
a PLate thickness: 12mm or 28mm 0,2 w In ) NOTCH FOR
NOTCH FOR FATIGUE BEND TEST
STATIC BEND TEST
FIG.1 TEST SPECIMENS FOR STATIC-BEND TEST
e -J 22 20 Notch. I Notch. 2 Notch. 3 15
(o
/1
2Position of notches in test-pieces
q2 0,4 0,6 0,8
C.0.D
(fracture)-1,2 1,4 1,6 1,8
FIG.2 C.0.D.- RECORD OF 12mm SPECIMENS.
12
36 Notch. 1 35
.-Notch.3
34I
Notch.2 33 -32I.
31 D-21,' __-°'
30 1. .RTD-/ 29..
29,.-/
/1
26 20le
16 14 12 Notch N! 3Postion of notches in testpieces
10
i4
2 I t I t I I I I I I -2
0,4 0,6 0,81,
1,2 1,4 1,6 1,8 2,-C.O.D. (fracture)L
TENSILE TEST
d
TP.- 6,7 and 8 IL-4 and 5
FIG.4 POSITION OF SPECIMEN
TABLE: TP- 1,2
'ir
E E SPECIMEN N2 - YIELD POINT kg/mm2 TENSILE STRENGTH kg/mm2 ELONGATION dp.5 (%) REMARKS TP- 1 36,2 52,8 33,5 PLATE MAT.TP-2
35,95 52,7 32,5 PLATE MAT.TP-3
36,2 52,8 33,2 PLATE MAT.TL-4
54,4 64,6 17,9 WELD MAT.TL-5
51,8 61,8 20,5 WELD MAT. TP-6 38,1 52,1 34,7 PLATE MAT. TP-7 37,7 52,2 34,- PLATE MAT. TP-8 37,5 52,4 34,,1 PLATE MAT. TPL-10 377 54,7 23,6 PLATE/WELD MAte
I
A-A
F:
.\ !Ir ¡1
i____ ___IU_ ____.i L...__. J hi
-a
POSITION OF CHARPV-V NOTCHES IN WELDED PLATE
CHARPY-V TEST RESULTS
}mm
¿'e'
Specimen N°
Notch position Test temp.
°C Charpy-V vaLues kgm/cm2 R.T.D.-1 PLate mat.