Utilization of Miniature C(T) specimens for the fracture toughness ...
Utilization of Miniature C(T) specimens for the fracture toughness evaluation of RPV steels by the Master Curve Method Materials Science Research Laboratory Central Research Institute of Electric Power Industry
Masato Yamamoto, Naoki Miura and Naoki Soneda EPRI International Light Water Reactor Materials Reliability Conference and Exhibition 2016
August 2016, Hyatt Regency McCormick Place, Chicago, IL, USA 1
Background Overall goal: Establish and standardize a test procedure to obtain T0 values using Mini-C(T) specimens machined from broken Charpy halves. If we can use Mini-C(T) specimens, which can be machined from broken Charpy specimens, additional fracture toughness data can be obtained from existing surveillance material. The Master Curve (MC) method can compensate the specimen size effect. It was confirmed that T0 values obtained using MiniC(T) specimens are consistent to those obtained by larger C(T) specimens up to 4T-C(T) specimens in three different heats. (ASME J. of PVT 134-021402) August, 2016
2
The Master Curve method o
T - T,1 0degC ( C) T-T 0 0
1 /2
KIc - 1 T (k s i·in) K Jc (1Teq), ksi√in
300 250 200 150 100
-1 5 0
-1 0 0
-5 0
0
50
1T ù MPa K IcJc(med 2 730 .3 ++ 6 370exp .6 e xp [0é 1 0 5 6 (T(T -T 1-T 0 0 )], ) . 1 /2 K (m e d ) ) == ë.00.019 o kûs,iｷin
H S S T -0 1 s u b a rc w e ld , N = 8 , T 1 0 0 = -1 0 4 .8 o F A 5 3 3 B c l. 1 s u b a rc w e ld , N = 8 , T 1 0 0 = -5 8 o F A 5 3 3 B c l. 1 w e ld , N = 1 0 , T 1 0 0 = -1 5 1 .6 o F A 5 3 3 B w e ld -H A Z , N = 6 , T 1 0 0 = -1 3 1 .8 o F A 5 0 8 c l.2 , N = 1 2 , T 1 0 0 = -1 1 9 .2 o F A 5 0 8 c l.2 , N = 9 , T 1 0 0 = -6 9 .8 o F A 5 0 8 c l.2 , N = 1 0 , T 1 0 0 = -5 .8 o F A 5 3 3 B c l.1 , N = 1 3 , T 1 0 0 = -7 4 .2 o F H S S T -0 1 , N = 1 7 , T 1 0 0 = -2 .2 o F H S S T -0 2 , N = 6 9 , T 1 0 0 = -1 8 .4 o F H S S T -0 3 , N = 9 , T 1 0 0 = 2 8 .4 o F
99% 97% 95%
350 300 250 200
50%
150 5% 3% 1%
50 0
100
1 /2 - 1 T ,(M P a ·m) KKJcIc (1Teq) Pa√m
350
50 0
-3 0 0
-2 0 0
-1 0 0
0
100
o
TT-T - T 1 0,0 degF ( F) 0
Fracture toughness (KJc) distribution can be described with specific Weibull distribution. Unique parameter To for indexing KJc distribution and its temperature trend. Specimen size dependency of fracture toughness can be compensated. August, 2016
3
Mini-C(T) specimen (
Dimension 4mm thickness C(T) specimen, whose principal dimensions (B, W, a0, L, 2H) are the same as those standardized in ASTM E1921. Specimen size (4 x 10 x 9.6 mm) Small enough to be machined maximum 4 specimens from a broken Charpy half. August, 2016
)
B (4±0.08)
2H (9.6±0.08)
2.2±0.04
3
N
2.2±0.04
D (f2±0.04)
0.8 at side surface a0 (4±0.4) W (8±0.04) L (10±0.08)
4
Development and standardization of Mini-C(T) test and evaluation technique Steps Basic applicability of MiniC(T) technique
Round robin testing
Standardization of testing and evaluation standard
Implementation of MiniC(T) technique in surveillance program
Un-irradiated RPV base metal
PVT2012, vol. 134, 021402 (2012)
ASTM STP1576, STP157620140020 (2015)
Revision of ASTM E1921 (2016) and JEAC4216(2015)
(2016~?)
Un-irrad. weld and HAZ
PVP2015-45545 PVP2016-63762 (2015, 2016)
Small RR among HZDR, SCK/CEN, EPRI, CRIEPI
Irrad. base metal
JAEA project (2015-) CRIEPI project (2015~2018)
Independent projects on various irradiated specimens?
Irrad. weld and HAZ
JAEA project?
Completed, In progress, Planning August, 2016
5
Objective Understand “the size effect” of Mini-C(T) for the practical application How many Mini-C(T) specimens will be needed to evaluate the reference temperature, T0? What test temperature will be the most appropriate selection? August, 2016
6
Procedure Carry out the Monte-Carlo simulation with assuming the Weibull fracture toughness distribution of the Master Curve method. 1. Generate fracture toughness (KJc) data set sampled from the Master Curve Weibull distribution. 2. Estimate To and check its accuracy to the true value. 3. Repeat above two steps up to 1000 times to obtain the statistics of To values. August, 2016
7
Sampling of fracture toughness data
The cumulative probability of KJc in the Master Curve Pf=1-exp{-[(KJc-20)/(K0-20)]4} KJc : Individual fracture toughness K0 : Parameter related to To
Two materials (plate and weld) are simulated To: -115 degC (SQV2A base metal, unirradiated) To: -77 degC (SQV2A weld metal, unirradiated)
August, 2016
Yamamoto et al, ASME PVP2015-45545, (2015)
Random number source: Mersenne
8
Specimen size effect on KJc(limit) Lower KJc(limit) for smaller specimens. KJc data exceeding KJc(limit) are “Invalid” and to be censored. Increasing number of invalid KJc will lead to Larger fraction of invalid T0 . Less accurate estimation of T0 . In order to minimize invalid KJc, possible test temperature window has to be narrower in Mini-C(T) than the larger ones. August, 2016
9
MC – KJc(limit) relationship before and after irradiation
E= 210-0.063(T-20), σy (unirrad) = 84 +271 exp {110 / (T +273.15)} Miura, N. and Soneda, N., ASME, Journal of Pressure Vessel Technology,Vol. 134, 021402, (2010)
ΔTo =0.70 Δσy Sokolov, M.A., and Nanstad, R. K., ASTM STP 1325,(1999) August, 2016
10
Analytical conditions Clarify the effect of KJc(limit) for 4mm-T specimen in To evaluation. Number of specimens: N=8, 9, 10, 11, 12, 13, 14, 16, 20 and 50 To true value: -115 (base metal), -77 (weld metal) degC Shift of To due to irradiation : ΔTo = 0, 100 degC To determination procedure : ASTM E1921-15 Specimen : Mini-C(T) (10x9.6x4 mm C(T) specimen) Number of trials : 1000
August, 2016
11
Distribution of estimated To 80% of To within 5 ˚C
70% of To within 10 ˚C
T-To=-30˚C
T-To= 0˚C 60% of To within 5 ˚C
Very few valid To Estimated To is 10˚C higher than true value
Selection of T is very important. August, 2016
12
Fraction of valid To to 1000 trials
Invalid due to low number of valid KJc
T-T0
Masato Yamamoto, Naoki Miura and Naoki Soneda EPRI International Light Water Reactor Materials Reliability Conference and Exhibition 2016
August 2016, Hyatt Regency McCormick Place, Chicago, IL, USA 1
Background Overall goal: Establish and standardize a test procedure to obtain T0 values using Mini-C(T) specimens machined from broken Charpy halves. If we can use Mini-C(T) specimens, which can be machined from broken Charpy specimens, additional fracture toughness data can be obtained from existing surveillance material. The Master Curve (MC) method can compensate the specimen size effect. It was confirmed that T0 values obtained using MiniC(T) specimens are consistent to those obtained by larger C(T) specimens up to 4T-C(T) specimens in three different heats. (ASME J. of PVT 134-021402) August, 2016
2
The Master Curve method o
T - T,1 0degC ( C) T-T 0 0
1 /2
KIc - 1 T (k s i·in) K Jc (1Teq), ksi√in
300 250 200 150 100
-1 5 0
-1 0 0
-5 0
0
50
1T ù MPa K IcJc(med 2 730 .3 ++ 6 370exp .6 e xp [0é 1 0 5 6 (T(T -T 1-T 0 0 )], ) . 1 /2 K (m e d ) ) == ë.00.019 o kûs,iｷin
H S S T -0 1 s u b a rc w e ld , N = 8 , T 1 0 0 = -1 0 4 .8 o F A 5 3 3 B c l. 1 s u b a rc w e ld , N = 8 , T 1 0 0 = -5 8 o F A 5 3 3 B c l. 1 w e ld , N = 1 0 , T 1 0 0 = -1 5 1 .6 o F A 5 3 3 B w e ld -H A Z , N = 6 , T 1 0 0 = -1 3 1 .8 o F A 5 0 8 c l.2 , N = 1 2 , T 1 0 0 = -1 1 9 .2 o F A 5 0 8 c l.2 , N = 9 , T 1 0 0 = -6 9 .8 o F A 5 0 8 c l.2 , N = 1 0 , T 1 0 0 = -5 .8 o F A 5 3 3 B c l.1 , N = 1 3 , T 1 0 0 = -7 4 .2 o F H S S T -0 1 , N = 1 7 , T 1 0 0 = -2 .2 o F H S S T -0 2 , N = 6 9 , T 1 0 0 = -1 8 .4 o F H S S T -0 3 , N = 9 , T 1 0 0 = 2 8 .4 o F
99% 97% 95%
350 300 250 200
50%
150 5% 3% 1%
50 0
100
1 /2 - 1 T ,(M P a ·m) KKJcIc (1Teq) Pa√m
350
50 0
-3 0 0
-2 0 0
-1 0 0
0
100
o
TT-T - T 1 0,0 degF ( F) 0
Fracture toughness (KJc) distribution can be described with specific Weibull distribution. Unique parameter To for indexing KJc distribution and its temperature trend. Specimen size dependency of fracture toughness can be compensated. August, 2016
3
Mini-C(T) specimen (
Dimension 4mm thickness C(T) specimen, whose principal dimensions (B, W, a0, L, 2H) are the same as those standardized in ASTM E1921. Specimen size (4 x 10 x 9.6 mm) Small enough to be machined maximum 4 specimens from a broken Charpy half. August, 2016
)
B (4±0.08)
2H (9.6±0.08)
2.2±0.04
3
N
2.2±0.04
D (f2±0.04)
0.8 at side surface a0 (4±0.4) W (8±0.04) L (10±0.08)
4
Development and standardization of Mini-C(T) test and evaluation technique Steps Basic applicability of MiniC(T) technique
Round robin testing
Standardization of testing and evaluation standard
Implementation of MiniC(T) technique in surveillance program
Un-irradiated RPV base metal
PVT2012, vol. 134, 021402 (2012)
ASTM STP1576, STP157620140020 (2015)
Revision of ASTM E1921 (2016) and JEAC4216(2015)
(2016~?)
Un-irrad. weld and HAZ
PVP2015-45545 PVP2016-63762 (2015, 2016)
Small RR among HZDR, SCK/CEN, EPRI, CRIEPI
Irrad. base metal
JAEA project (2015-) CRIEPI project (2015~2018)
Independent projects on various irradiated specimens?
Irrad. weld and HAZ
JAEA project?
Completed, In progress, Planning August, 2016
5
Objective Understand “the size effect” of Mini-C(T) for the practical application How many Mini-C(T) specimens will be needed to evaluate the reference temperature, T0? What test temperature will be the most appropriate selection? August, 2016
6
Procedure Carry out the Monte-Carlo simulation with assuming the Weibull fracture toughness distribution of the Master Curve method. 1. Generate fracture toughness (KJc) data set sampled from the Master Curve Weibull distribution. 2. Estimate To and check its accuracy to the true value. 3. Repeat above two steps up to 1000 times to obtain the statistics of To values. August, 2016
7
Sampling of fracture toughness data
The cumulative probability of KJc in the Master Curve Pf=1-exp{-[(KJc-20)/(K0-20)]4} KJc : Individual fracture toughness K0 : Parameter related to To
Two materials (plate and weld) are simulated To: -115 degC (SQV2A base metal, unirradiated) To: -77 degC (SQV2A weld metal, unirradiated)
August, 2016
Yamamoto et al, ASME PVP2015-45545, (2015)
Random number source: Mersenne
8
Specimen size effect on KJc(limit) Lower KJc(limit) for smaller specimens. KJc data exceeding KJc(limit) are “Invalid” and to be censored. Increasing number of invalid KJc will lead to Larger fraction of invalid T0 . Less accurate estimation of T0 . In order to minimize invalid KJc, possible test temperature window has to be narrower in Mini-C(T) than the larger ones. August, 2016
9
MC – KJc(limit) relationship before and after irradiation
E= 210-0.063(T-20), σy (unirrad) = 84 +271 exp {110 / (T +273.15)} Miura, N. and Soneda, N., ASME, Journal of Pressure Vessel Technology,Vol. 134, 021402, (2010)
ΔTo =0.70 Δσy Sokolov, M.A., and Nanstad, R. K., ASTM STP 1325,(1999) August, 2016
10
Analytical conditions Clarify the effect of KJc(limit) for 4mm-T specimen in To evaluation. Number of specimens: N=8, 9, 10, 11, 12, 13, 14, 16, 20 and 50 To true value: -115 (base metal), -77 (weld metal) degC Shift of To due to irradiation : ΔTo = 0, 100 degC To determination procedure : ASTM E1921-15 Specimen : Mini-C(T) (10x9.6x4 mm C(T) specimen) Number of trials : 1000
August, 2016
11
Distribution of estimated To 80% of To within 5 ˚C
70% of To within 10 ˚C
T-To=-30˚C
T-To= 0˚C 60% of To within 5 ˚C
Very few valid To Estimated To is 10˚C higher than true value
Selection of T is very important. August, 2016
12
Fraction of valid To to 1000 trials
Invalid due to low number of valid KJc
T-T0
