onset of fatigue cracking in through wall sg flaws - an
TRANSCRIPT
O t f F ti C ki iOnset of Fatigue Cracking in Through Wall SG Flaws
An Evaluation of Time Dependent Leak Rates
Jim Begley – TCA Solutions
1
An Evaluation of Time Dependent Leak Ratesp
Argonne National Labs has performed leak rate tests of alloy 600 tubing with lab-grown axial ODSCC flaws and EDM notchesg g
– In some cases dramatic increases in leak rates were observed under nominally constant pressure conditions.
SGMP f d d j t t d t i th ig ifi f thi SGMP funded a project to determine the significance of this ANL leak rate test data to the leakage integrity of operating SGs. – Fractographic studies of some previous leak rate specimens were
performed.
– Additional leak rate tests of specimens with sharp fatigue cracks were performed in the ANL High Pressure Blowdown Facility.
2
An Evaluation of Time Dependent Leak Rates
Time dependent increases in leakage of operating steam generator tubes containing stress corrosion cracks has been observed in the
t d i t t d th g d l f il f ll lig t past and interpreted as the gradual failure of small ligaments between crack segments.
Initial ANL leak rate tests raised the possibility of leakage induced fatigue crack growth leading to large increases in leak rates.
3
An Evaluation of Time Dependent Leak RatesAn Evaluation of Time Dependent Leak Rates
Five types of crack growth have been identified in ANL leak rate tests: – Small scale, time independent mechanical tearing of
tapered depth ligaments as internal pressure is increasedS f i f i f i f– Small scale low cycle fatigue failure of small ligaments from stop/start pressurization cycles
– Small scale, time dependent fracture of small ligaments via time dependent plastic deformation at very high time dependent plastic deformation at very high stress/strain levels
– Small scale, time dependent fatigue cracking of tapered depth ligaments via pump operation and bypass flow effects p g p p p ypin the room temperature, high flow pump test system
– Large scale, time dependent fatigue cracking from fluid jet/structure interactions at high flow rates (?)
9
An Evaluation of Time Dependent Leak RatesAn Evaluation of Time Dependent Leak Rates
ANL Room Temperature High Flow Pump ANL Room Temperature High Flow Pump
10
An Evaluation of Time Dependent Leak RatesAn Evaluation of Time Dependent Leak Rates
ANL Room Temperature High Flow Pump ANL Room Temperature High Flow Pump
11
An Evaluation of Time Dependent Leak RatesAn Evaluation of Time Dependent Leak Rates
ANL Room Temperature High Flow Pump ANL Room Temperature High Flow Pump
12
An Evaluation of Time Dependent Leak RatesAn Evaluation of Time Dependent Leak Rates
SCC
Crack Grow thDuring LeakCrack OpeningAfter Leak
SGL-750
Axial ODSCC
13
An Evaluation of Time Dependent Leak RatesAn Evaluation of Time Dependent Leak Rates
SGL-750
Axial ODSCC
14
An Evaluation of Time Dependent Leak RatesAn Evaluation of Time Dependent Leak Rates
SGL-750
Axial ODSCC
15
An Evaluation of Time Dependent Leak RatesAn Evaluation of Time Dependent Leak Rates
SGL-750
Axial ODSCC
16
An Evaluation of Time Dependent Leak RatesAn Evaluation of Time Dependent Leak Rates
NUANOD.2DED M N otch Zone
Trapezoidal EDM NotchCrack G row th During Leak Test
Crack O pening After Leak Tes t
Tube ODTube ID
17
Onset of Fatigue Cracking in Through Wall SG FlawsOnset of Fatigue Cracking in Through Wall SG Flaws
To perform a leak rate test under absolute t t diti i constant pressure condition, i.e. no pressure
oscillation by a pump, it was decided to use Blowdown Test Facility. Refurbishment and upgrade of the Blowdown Test Facility was completed in 2010, which was funded by SGMP and NRCfunded by SGMP and NRC.Leak tests were performed in 2011 at room temperature at a constant pressure of 2500 psig using rectangular axial fatigue crack specimens. – The time of interest is 1 to 24 hours and the leak
rate of interest is 0.05 to 1.5 gpm.
18
rate of interest is 0.05 to 1.5 gpm.
Onset of Fatigue Cracking in Through Wall SG FlawsOnset of Fatigue Cracking in Through Wall SG Flaws
Alloy 600 tube specimen is surrounded by six neighboring tubes.T b t l t h ldi g th ighb i g Tube support plates are holding the neighboring tubes.
19
Onset of Fatigue Cracking in Through Wall SG Flaws
Rectangular axial fatigue crack was grown from an EDM starter notch on 3/4” OD Alloy 600 tube specimen
Hydraulic cylinder generates the pressure oscillation
20
– Hydraulic cylinder generates the pressure oscillation
Onset of Fatigue Cracking in Through Wall SG Flaws Preliminary Test SpecimenPreliminary Test Specimen
Practice to make sure the system worked0.68 in-long axial through wall fatigue crack, which was grown from a 0.5 in-long & from a 0.5 in long & 80%TW EDM starter notch.The flow rate increased f 5 45 t 5 65 g from 5.45 to 5.65 gpm during 30 min leak testing at 2,350 (±7) psig.Fractography indicated crack tip ductile stretching not fatigue growth
21
growth
Onset of Fatigue Cracking in Through Wall SG Flaws First Test SpecimenFirst Test Specimen
0 38 in-long axial through wall fatigue crack which was 0.38 in-long axial through wall fatigue crack, which was grown from 0.25-in long & 40%TW axial EDM starter notch.Leak rate of 0 35 gpm
22
Leak rate of ~0.35 gpm.
Onset of Fatigue Cracking in Through Wall SG FlawsSecond Test SpecimenSecond Test Specimen
0.47 in-long axial through wall fatigue crack, which was grown from 0.25-in long & 40%TW axial EDM starter notch.
L k t f 0 9 g24
Leak rate of ~0.9 gpm.
Onset of Fatigue Cracking in Through Wall SG FlawsThird Test SpecimenThird Test Specimen
0 52 in long axial through wall fatigue crack which 0.52 in-long axial through wall fatigue crack, which was grown from 0.25-in long & 40%TW axial EDM starter notch.
26Leak rate of ~2 gpm.
Onset of Fatigue Cracking in Through Wall SG FlawsFourth Test SpecimenFourth Test Specimen
0.68 in-long axial fatigue crack, which was grown from 0 25-in long & 40%TW axial EDM starter notch0.25-in long & 40%TW axial EDM starter notch.
Leak rate of ~8 gpm @ 2350 psig and ~11 gpm @ 2500 psig.
28
p g
Onset of Fatigue Cracking in Through Wall SG FlawsFourth Test SpecimenFourth Test Specimen
Due to the limitation of a turbine flow meter (max. 8 gpm) the pressure was set to 2350 psig.
29
p p gGradual increase from ~7 to 7.8 gpm was observed.
Onset of Fatigue Cracking in Through Wall SG FlawsFourth Test Specimen (2nd & 3rd Tests)Fourth Test Specimen (2 & 3 Tests)
During the 2nd test the leak rate still showed slight increase. However, from the 3rd testing the leak rate became stable at ~8 gpm.
During the accumulated test time of 2 hr the leak rate did not increase at During the accumulated test time of 2 hr the leak rate did not increase at the constant pressure.
The leak rate estimated from the weight change appears to agree with that from the turbine flow meter except the initial 5 min suggesting that data for
30
from the turbine flow meter, except the initial 5 min, suggesting that data for the initial 5 min may not be close to actual flow rate.
Onset of Fatigue Cracking in Through Wall SG FlawsFourth Test Specimen (4th 5th & 6th Tests)Fourth Test Specimen (4 , 5 & 6th Tests)
31
Onset of Fatigue Cracking in Through Wall SG Flaws
No fatigue crack growth even
SCC
No fatigue crack growth, even of small ligaments !
SCC
Non-SCC Ligament 2
Ligament 1
34
Flow Induced Fatigue Cracking IssueL k T t C l iLeak Test Conclusions
Test results indicate that for operating steam generator conditions there is no issue of flow induced fatigue crack growth leading to large increases in leak rates.
Th light i i l k t i th l t t i f t t i The slight increases in leak rates in the latest series of test is caused by time dependent crack tip plasticity, not fatigue growth.
The small observed increases of leak rates over time does raise the issue of required hold times for in situ leak testing.
35
Time Dependent Plasticity in Alloy 600
At very high stresses primary stage creep is observed for many structural alloys at room temperature.Creep testing at high stress levels has been performed for Alloy 600 at room temperature.Wire specimens (0.032” diameter) of mill annealed Alloy 600 have been subjected to dead weight loading at room temperature to obtain strain versus time curves.Thi i f ti h b d t l t th d f This information has been used to evaluate the adequacy of a 5 minute hold time for measurement of leak rates during in situ testing.A small increase in test pressure is equivalent to very much A small increase in test pressure is equivalent to very much longer hold times.
36
Time Dependent Plasticity in Alloy 600
Room Temperature100000
70000
80000
90000
s
40000
50000
60000
gine
erin
g St
ress
10000
20000
30000
En
00 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4
Engineering Strain
37
Time Dependent Plasticity in Alloy 600
Room Temperature140000
100000
120000
60000
80000
True
Str
ess
20000
40000
00 0.05 0.1 0.15 0.2 0.25 0.3
True Strain
38
Time Dependent Plasticity in Alloy 600
0.25
0 15
0.2
, in.
/in. 83.9 ksi
0.1
0.15
gine
erin
g St
rain
,
71.1 ksi
77.5 ksi
0.05
En
51.9 ksi
58.3 ksi
64.7 ksi
00 2000 4000 6000 8000 10000 12000 14000 16000 18000
Time, seconds
45.4 ksi
39
Time Dependent Plasticity in Alloy 600
0.035
0.025
0.03
in./i
n.
77.5 ksi
0.015
0.02
crea
se in
Str
ain,
45.4 ksi
0.005
0.01
Inc
00 50 100 150 200 250 300 350
Time, seconds
40
Time Dependent Plasticity in Alloy 600
0.3592.0 ksi
0.25
0.3
n/in
. 88.8 ksi
90.4 ksi
0.15
0.2
inee
ring
Stra
in, i
n
83.9 ksi85.6 ksi
87.1 ksi
0.05
0.1
Eng
00 2000 4000 6000 8000 10000 12000 14000 16000 18000 20000
Time, seconds.
41
Time Dependent Plasticity in Alloy 600
0 28
0.32
90.4 ksi
0 2
0.24
0.28
./in.
0.12
0.16
0.2
ngin
eeri
ng S
trai
n, in
.
0.04
0.08
En
00 2000 4000 6000 8000 10000 12000 14000 16000 18000 20000
Time,seconds
42
Time Dependent Plasticity in Alloy 600
1000000
10000
100000
ds
1000
Tim
e to
Fai
lure
, sec
on
10
100
190000 90500 91000 91500 92000 92500 93000 93500 94000
Stress, psi
43
Adjustment of In Situ Leakage Test TimeSummary/Conclusiony/
Time dependent plasticity in Alloy 600 is equivalent to about a 2% decrease in the flow strength.gThis increases leak rates by about 10%.A 2% increase in the test pressure for in situ pleakage testing is equivalent to increasing the test time to several hours.The existing conservatisms (15%) in the temperature correction factor between room and operating temperature flow strength and the uncertainty temperature flow strength and the uncertainty factor of the pressure measurement is sufficient to account for the small level of time dependent
44
increases in leak rate.
ConclusionsConclusions
Leak test results indicate that for operating Leak test results indicate that for operating steam generator conditions there is no issue of flow induced fatigue crack growth leading o o duced at gue c ac g o t ead gto large increases in leak rates.
Existing conservatisms are sufficient to Existing conservatisms are sufficient to account for the possibility of small time dependent increases of leak rate.dependent increases of leak rate.
SGTF recommends that this issue be closed.
45