y = - 0.79281 + 0.27997~ R"2 = 0.984
2.0 -
1.5 -
1.0 -
0.5 -
0.0 ' I I I I
2 4 6 8 1 0 1 2
NUMBER OF STRESSCOAT CRACKING, (Line/mm)
Fig. D3-7-Relationship of Stresscoat cracking intensity to material strain
Causes and Repair of Cracking 91
Copyright American Petroleum Institute Reproduced by IHS under license with API
Not for Resale No reproduction or networking permitted without license from IHS
--`,,,,`,-`-`,,`,,`,`,,`---
~~
~
A P I PUBL*938 76 M 0 7 3 2 2 9 0 0560441 109
O AVE. CRACK DEPTH TEST TEMPERATURE: 621OC (1 150OF)
in.) in deflection in the test fixture and the number of cracks across a 10 mm length adjacent to the weldtoes were determined. The linear crack density was deter- mined to be approximately 4.7/mm which is equiva- lent to 0.53% strain.
Considering only elastic deformation in the loaded PREVEW test sample, a strain of 0.53% can be calcu- lated to be equivalent to a stress of approximately 1096 MPa (159 Ksi). Since the yield strength of l%Cr-l/zMo steel is approximately 414 MPa (60 Ksi), an elastic-plastic theory must be used to calculate the resultant stress for 0.53% applied strain. The stress adjacent to the weld toe in the fillet joint is calculated to be approximately 483 MPa (70 Ksi) for 0.53%
strain which corresponds to a deflection of 0.3 mm (0.12 in.). However, it must be recognized that the hardness or strength of the CGHAZ is significantly higher than that of the base plate. Thus, the stress in the CGHAZ region at the toe of the weld may be higher than 70 Ksi (considering most of the addi- tional strain is relaxed by plastic deformation of the base plate).
Test Criteria
The following criteria were used t o assess suscepti- bility to reheat cracking in the PREVEW test: maxi- mum crack depth, average crack depth and crack length. The crack depth is measured in three metallo- graphically-prepared transverse sections of both fillet welds or the butt joint (depending on the test configu- ration), at 1 0 0 ~ to 4 0 0 ~ . The average of the crack
depth measured in the three transverse sections is used as a criterion.
Average crack depth (mm) = (Sum of crack/depth)/3 The maximum crack depth in three transverse sections can also be used as a criterion for assessment of reheat cracking susceptibility of a material using the PREVEW test.
The crack length criterion is defined as the percent- age of crack length to the total weld bead length in two longitudinal sections along the fillet joint. The crack length is measured at 100 x to 400 x using an optical microscope
Crack length (Or,)
100 x Sum of crack lengths at both toes 2 x Weld bead length
- -
Test Time and Reheat Cracking Susceptibility It has been suggested that reheat cracking is dependent on the time and temperature of the PWHT cycle. During PWHT or subsequent elevated tempera- ture service, the welding residual stresses relax in a temperature range which also corresponds to the temperature range of reheat cracking susceptibility.
In this study, cracking initiation and propagation behavior was investigated by examining the crack depth' in three transverse sections and crack length in two longitudinal sections of the fillet weld samples of UT3 tested at 621°C (1150°F) for various test times (Figs. D3-8, D3-9). It was found that heating to 621°C
I+ HEATING TO 621°C (1 150°F)
n
E E
n n
a
u
I I- W Y o
K o
3
O MAX.CRACK
' II DEPTH
O 1 0 20 30 40
TEST TIME (hours)
Fig. D3-%Correlation of reheat crack depth with test time in fillet weld of UT3 using PREVEW test
92 Causes and Repair of Cracking
Copyright American Petroleum Institute
--`,,,,`,-`-`,,`,,`,`,,`---
A P I PUBLU938 9b œ 0732290 O560442 O45 œ
1 O0
60
40
20 -
O +
n s
I l-
13 z
u1 -I
Y o
K o
Y
a
HEATING TO 621°C (1150OF)
TEST TEMPERATURE: 621 "C (1 150°F)
8
;
;
I I I
O 1 0 2 0 30 40
TEST TIME (hours)
Fig. DB-%-Correlation of reheat crack length with test time in fillet weld of UT3 using PREVEW test
(1150°F) from room temperature in 1 hour, reheat cracking had already occurred in the PREVEW sample with a maximum crack depth of 2 mm and a crack length of 43%. With an increase in the test time the cracks propagated. Figure D3-8 shows that the exten- sion of the reheat cracks does not occur beyond approximately 4 hours at 621°C (1150"F), which illustrates that most of the applied stress relaxes in approximately 4 hours.
A photograph of a dye penetrant inspection of a reheat cracked PREVEW specimen is shown in Fig.
D3-10. Weld toe cracks along both weld beads are
clearly revealed. The crack depth and crack length observed in the transverse and longitudinal sections of the fillet joint are shown in Figs. D3-11 and D3-12, respectively. It is clear that reheat cracks initiate at the weld toes and then propagate along the grain boundaries in CGHAZ, which is a typical reheat cracking morphology. Figure D3-13 shows a trans- verse section of the butt joint of UT3 with notch in CGHAZ tested at 621°C (1150°F) for 30 hours. It is clear that the crack initiated at the tip of notch and propagated approximately 2.5 mm along the grain boundaries in CGHAZ.
Fig. D 3 - l G D y e penetrant inspection of reheat cracking in fillet weld
Causes and Repair of Cracking 93
Copyright American Petroleum Institute Reproduced by IHS under license with API
Not for Resale No reproduction or networking permitted without license from IHS
--`,,,,`,-`-`,,`,,`,`,,`---
API P U B L * ù 3 8 96 m 3 7 3 2 2 9 0 0560443 T B 1
a. Transverse section 3.2X
Fig. D3-13a-Transverse section of butt weld in UT3.4.6x
b. Longitudinal section 1.8X
Fig. 03-1 1-Reheat cracking in fillet weld of UT3 tested at 1150°F for 30 hours
a. Transverse section 22.5X
b. Longitudinal section 50X
Fig. D3-12-Reheat cracking initiated at weld toes in HAZ of UT3
Fig. D3-13b-Reheat crack initiated at the notch tip. Propagated along CGHAZ of butt weld in UT3,50x
94 Causes and Repair of Cracking
Copyright American Petroleum Institute
--`,,,,`,-`-`,,`,,`,`,,`---
A P I P U B L * 9 3 8 9 6 0 7 3 2 2 9 0 O 5 6 0 4 4 4 9 1 8
Fig. D3-13c-Intergranular reheat crack in CGHAZ in butt weld of UT3,200 x
Correlation of PREVEW Test Results and Gleeble Rupture Test Results
The PREVEW test Results for samples UT2 thru UT17 are given in Table D3-3.
All of the heats tested in the PREVEW test were also investigated utilizing the Gleeble stress rupture test (described in Appendix D-1 of this report). The PREVEW test crack depth and length is plotted against the Gleeble reduction in area in Figs. D3-14, D3-15 and D3-16. If the Gleeble reduction in area is less than approximately 13% cracking occurs in the PREVEW test.
Vinckier and Pense12 studied 18 steels using simu- lated specimens tensile tested at elevated tempera-
ture. A criterion for cracking based on the reduction in area of these materials previously subjected to simulated welding thermal cycles was established.
The reheat cracking susceptibility was ranked as:
Susceptibility to Reduction reheat cracking in area (%) Extremely susceptible < 5 Not susceptible > 20 Highly susceptible 5-10 Slightly susceptible 10-15
The authors stated that any material capable of 20%
RA in the CGHAZ at the specified PWHT tempera- ture would not exhibit reheat cracking. The PREVEW test results are in agreement with this conclusion.
However, the 20% RA as a criterion appears to beconservative based on the PREVEW reheat crack- ing test results.
Conclusions
The new reheat cracking test, the PREVEW test, has the advantages of both self-restrained weld tests and simulated weld tests. The PREVEW test can be applied in both fillet and butt weld configurations to typify closely the actual fabrication situations. The test conditions, such as welding parameters, restraint stress and test temperature, can be chosen identical to actual fabrication conditions.
The crack depth and length can be used as criteria to evaluate reheat cracking susceptibility of Cr-Mo steels. The test temperature and test time should be determined based on the PWHT or the in-service conditions.
The Stresscoat technique was successfully applied to determine the stress in the PREVEW test sample.
The stress at the weld toes in fillet joint is about 483 MPa (70 Ksi) when the sample is loaded to 0.12 in. in deflection in severe restraint condition.
The PREVEW test results are satisfactorily repro- ducible and are in accordance with Gleeble reheat cracking test results. If the reduction in area of Gleeble stress rupture test is over 13% no reheat cracking occurs in the PREVEW test.
Causes and Repair of Cracking 95
Copyright American Petroleum Institute Reproduced by IHS under license with API
Not for Resale No reproduction or networking permitted without license from IHS
--`,,,,`,-`-`,,`,,`,`,,`---
~ ~
A P I PUBLa938 ù b 0732290 0 5 6 0 4 4 5 8 5 4
N
7
+I
? 5
m I- 3
96 Causes and Repair of Cracking
Copyright American Petroleum Institute
--`,,,,`,-`-`,,`,,`,`,,`---
A P I P U B L * ù 3 ụ 96 0732290 0560446 790
c a
r-x Q
.c. c
O 3
.-
Causes and Repair of Cracking 97
Copyright American Petroleum Institute Reproduced by IHS under license with API
Not for Resale No reproduction or networking permitted without license from IHS
--`,,,,`,-`-`,,`,,`,`,,`---
A P I PUBL*938 9 b 0732290 0 5 6 0 4 4 7 6 2 7 W
c
-- O
.- o c 9 z
n
Y- a
c a, Q
P c
c
O 7
.-
v) a, Q S
-
2 I
o o
5 al rn t O
a
U 3
v) a
a,
-
L
E 5 Q U a
Y o
O
O
a
3 -r O
c
.- L L
L
T- L
\c O
0. O
m m
a
U a,
7
+
F
E a,
a o
ru
3 +
E
+
Ia 3
c.
L
a,
O
c +
c c
al
0
a,
2
ó
c c c
98 Causes and Repair of Cracking
Copyright American Petroleum Institute
--`,,,,`,-`-`,,`,,`,`,,`---
A P I P U B L r ù 3 8 96 0732290 0560448 563 m