Development of New Design Fatigue Curves in Japan · 2018. 9. 25. · Development of New Design Fatigue Curves in Japan ... data but also fatigue data of mean stress effect. •Fatigue

0© 2016 MITSUBISHI HEAVY INDUSTRIES, LTD. All Rights Reserved.© 2018 MITSUBISHI HEAVY INDUSTRIES, LTD. All Rights Reserved.

Seiji Asada Yun Wang

Mitsubishi Heavy Industries, Ltd. Hitachi, Ltd.

Masahiro Takanashi Kentaro Hayashi

IHI Corporation The Kansai Electric Power Co., Inc.

Development of New Design Fatigue Curves in Japan

- Discussion of Best Fit Curves based on Fatigue Test Data -

Environmentally Assisted Fatigue (EAF) Research and Related ASME Activities,

NRC Public Meeting, September 25, 2018

1© 2016 MITSUBISHI HEAVY INDUSTRIES, LTD. All Rights Reserved.© 2018 MITSUBISHI HEAVY INDUSTRIES, LTD. All Rights Reserved. 1

Outlines

• Introduction

• Fatigue Tests Using Small Specimens [*1]

• Large-Scale Fatigue Tests Using Carbon and Low-Alloy Steel

Plates [*2]

• Large-Scale Fatigue Tests Using Stainless Steel Piping [*3]

• Conclusions

(Notes)

The details of the above fatigue experimental tests are shown in the following

2018 PVP papers.

[*1] Wang, Yun, et al., “Development of New Design Fatigue Curves in Japan -

Discussion of Best-Fit Curves Based on Fatigue Test Data With Small-Scale

Test Specimen -,” PVP2018-84052, ASME, 2018.

[*2] Takanashi, M., et al., “Development of New Design Fatigue Curves in Japan

-Discussion of Best-Fit Curves Based on Large-Scale Fatigue Tests of

Carbon and Low-Alloy Steel Plates -,” PVP2018-84456, ASME, 2018.

[*3] Bodai, M., “Development of New Design Fatigue Curves in Japan -

Discussion of Best-Fit Curves Based on Fatigue Test Data with Large-Scale

Piping -,” PVP2018-84436, ASME, 2018.


Introduction

• The DFC1/DFC2 subcommittee has not only developed a new fatigue evaluation method but also produced beneficial outcomes. To support this study, a Japanese utility project performed not only large scale fatigue tests using carbon & low-alloy steel flat plates and austenitic stainless steel piping but also fatigue tests using small specimens to obtain not only basic data but also fatigue data of mean stress effect.

• Fatigue life of a small specimen is generally defined as the number of cycles by 25% load drop, and this is considered to correspond to 3mm-deep crack in the test specimen.

• Hence, the fatigue lives of the large-scale fatigue tests are compared with the best-fit curve developed by the DFC1 subcommittee and the fatigue lives obtained by the small specimen fatigue tests.

• In this presentation, the fatigue tests using small specimens and large scale fatigue tests using carbon & low-alloy steel flat plates and austenitic stainless steel piping are summarized.


Fatigue Tests Using Small Specimens [*1]

• Materials

Material C Si Mn P S Ni Cr Mo Cu Fe

SUS316LTP 0.012 0.44 1.76 0.024 0.000 14.47 17.38 2.62 － Bal.

STPT370 0.200 0.25 0.82 0.014 0.001 －－－－ Bal.

SQV2A 0.18 0.24 1.43 0.005 0.002 0.66 0.11 0.520 0.001 Bal.

SCM435H 0.37 0.28 0.76 0.014 0.011 0.08 0.91 0.15 － Bal.

Material su (MPa) s0.2 (MPa)Elongation

(%)

Reduction of

Area (%)

SUS316LTP 556 (≥480) 238 (≥175) 53 (≥35) 86

STPT370 493 (≥370) 272 (≥215) 32 (≥30) 68

SQV2A 597 (550-690) 450 (≥345) 26 (≥18) 77

SCM435H 1074 (≥930) 991 (≥785) 17 (≥15) 58

[Notes]

- SUS316LTP (≈SA312 TP316L) was taken from the large-scale piping.

- STPT370 (≈SA106) is a carbon steel piping.

- SQV2A (≈SA533 Gr.B Cl.1) was taken from the large-scale Low-Alloy Steel (LAS) plate.

- SCM435H is a Cr-Mo steel with high tensile strength.

[wt%]


Fatigue Tests Using Small Specimens [*1] (continued)

• Fully Reversed Axial Fatigue Tests

SUS316LTP STPT370

BFC of DFC Subcommittee (σu = 556 MPa) BFC of DFC Subcommittee (σu = 493 MPa)

𝑺𝒂 = 𝟏. 𝟐 × 𝟏𝟎𝟓 − 𝟐𝟖𝝈𝒖 𝑵𝒇−𝟎.𝟓𝟖 + 𝟎. 𝟒𝟓𝝈𝒖 + 𝟑𝟔𝑺𝒂 = 𝟓. 𝟎𝟗 × 𝟏𝟎𝟒 𝑵𝒇

−𝟎.𝟒𝟖𝟓 + 𝟎. 𝟒𝟖𝟖𝝈𝒖

[Best-Fit Curve of DFC Subcommittee]

For Stainless Steels


For CS&LAS Steels



• Fully Reversed Axial Fatigue Tests (continued)

SQV2A SCM435H

BFC of DFC Subcommittee (σu = 597 MPa) BFC of DFC Subcommittee (σu = 1074 MPa)

𝑺𝒂 = 𝟏. 𝟐 × 𝟏𝟎𝟓 − 𝟐𝟖𝝈𝒖 𝑵𝒇−𝟎.𝟓𝟖 + 𝟎. 𝟒𝟓𝝈𝒖 + 𝟑𝟔


For CS&LAS Steels



• Fully Reversed Axial Fatigue Tests (continued)

SUS316LTP (εta=0.15%, Nf=1.06x106 cycles) STPT370 (εta=0.135%, Nf=5.40x105 cycles)

SQV2A (εta=0.15%, Nf=1.89x107 cycles) SCM435H (εta=0.3%, Nf=1.01x105 cycles)



• Mean Stress Correction

✓Modified Goodman Approach

𝜎𝑒𝑞 =𝜎𝑢𝜎𝑎

𝜎𝑢 − 𝜎𝑚=

𝜎𝑎1 − Τ𝜎𝑚 𝜎𝑢

✓Gerber Approach

𝜎𝑒𝑞 =𝜎𝑎

1 − Τ𝜎𝑚 𝜎𝑢𝟐

✓Peterson Approach

𝜎𝑒𝑞 =7𝜎𝑎

8 − 1 + Τ𝜎𝑚 𝜎𝑢𝟑

✓Smith-Watson-Topper Approach

𝜎𝑒𝑞 = 𝜎𝑎 ∙ 𝜎𝑚𝑎𝑥 = 𝜎𝑎 ∙ 𝜎𝑎 + 𝜎𝑚

𝜎𝑎: Stress Amplitude, 𝜎𝑒𝑞: Equivalent Stress Amplitude,

𝜎𝑚: Mean Stress, 𝜎𝑚𝑎𝑥: Maximum Stress Amplitude, 𝜎𝑢: Tensile Strength

Mean Stress (MPa)

0 200 400 600 800 1000 12000

200

400

600

800

σu =1000 MPa

σy =0.72 σu

σw0 =0.45 σu

Gerber

Peterson

S-W-T

Mod. Goodman

σy − σy

Fati

gu

e E

nd

ura

nce L

imit

(

MP

a)



• Mean Stress Correction

SUS316LTP STPT370

SCM435HSQV2A

Conservative


Large-Scale Fatigue Tests Using CS&LAS Plates [*2]

• Test Specimen and Test Machine

Stress Concentration Factor = 1.27


Large-Scale Fatigue Tests Using CS&LAS Plates [*2](continued)

• Test Results

ID Mat. AimUpper strain

(%)

Strain amp.

(%)

Mean strain

(%)

CS1 CS Size effect 0.24 0.24 0

CS2 CS Mean stress 0.3 0.24 0.06

LAS1 LAS Size effect 0.22 0.22 0

LAS2 LAS Size effect 0.18 0.18 0

LAS3 LAS Mean stress 0.3 0.22 0.08

LAS4 LAS Mean stress 0.3 0.18 0.12

[*] Fatigue Life: Crack penetrated

the plate width or the load

decreased


Large-Scale Fatigue Tests Using Stainless Steel Piping[*3]


Notched Portion

≈3,000

ϕ170.3

ϕ216.3

ϕ216.3233

≈40.5

≈50.6

Point A

(Center)

Point B

(Edge)XLoad

[Load Cell]

Strain Gages

Displacement, δa

[Actuator]

u1 u2 u4 u5

d1 d2 d4 d5d3

Thermocouple

L

u8

u10 – u14

u9

u6 u7

Stress Concentration Factor = 1.39


Large-Scale Fatigue Tests Using SS Piping[*3](continued)


Strain

AmplitudeMean Strain

TP-A ±0.44% No

TP-B ±0.44% No

TP-C ±0.25% No

TP-D ±0.25% +2.25%

TP-E ±0.5% +2.0%

TP-F ±0.2% +2.3%



✓ The fatigue lives of pipes for 3 mm crack and through-wall crack (TWC) are compared with the fatigue lives of the small specimens and the estimated best fit curve developed by the DFC subcommittee.

0.1

1.0

10.0

1.E+02 1.E+03 1.E+04 1.E+05 1.E+06 1.E+07

ひずみ範囲（

1/2

No

r1

/2N

25

）⊿

ε（％）

繰返し数 N（cycle）

材質：SUS316L

環境：室温大気中

○：大型配管（平均応力なし）（き裂深さ3㎜）

●：大型配管（平均応力なし）（貫通20㎜）

△：大型配管（平均応力あり）（き裂深さ3㎜）

▲：大型配管（平均応力あり）（貫通20㎜）

□：小型試験片（ N25 ）

－：TS＝542MPa推定線

試験F（◆）

貫通20㎜は推定値

（試験A～Eの20㎜（貫通）き裂に対

する3㎜き裂の繰返し数比の平均

0.868から推定）平均

試験A（○、●）

ひずみ範囲：0.98％、1.00％

Number of Cycles

Str

ain

, R

an

ge a

t 1/2

No

r 1/2

N2

5, ∆ε

(%)

SUS316LTP (in air)

: Pipe [No Mean Stress, 3mm crack]

: Pipe [No Mean Stress, TWC]

: Pipe [Mean Stress, 3mm crack]

: Pipe [Mean Stress, TWC]

: Small Specimens (N25)

--- : Estimated Best Fit Curve (TS=542MPa)

TP-A ( , )

∆ε: 0.98%, 1.00%

TP-F of Through Wall Crack( ):

[estimated from the average of

ratio between 3mm crack and

TWC for TP-A to TP-E]



[Fatigue Test] Target strain amplitude = ±0.44%

✓ Observation on Notched Portion by replica printing

Number of Cycles: 10,000 Number of Cycles: 11,450

Number of Cycles: 8,000 Number of Cycles: 9,000

Cracks



[Fatigue Test] Target strain amplitude = ±0.44% (continued)

✓ Observation of Fracture Surface (beach marking)

1

3㎜

23

3 4 56

7

8



• The data of the tested pipes are plotted for nominal strain amplitude and nominal mean stress calculated from the loads at 3 mm crack depth with mean stress correction.

• The S-W-T approach is more appropriate than the Modified Goodman approach

: TP-A (0%)

: TP-B (0%)

: TP-C (0%)

: TP-D (2.25%)

: TP-E (2%)

: TP-F (2.3%)

[Note] ( )= Mean Strain

: S-W-T

: Modified Goodman

Nominal Mean Stress

No

min

al

Str

ess A

mp

litu

de, σ

a(M

Pa)


Conclusions

• To support the new fatigue evaluation method by DFC1/DFC2 subcommittee, a Japanese utility project performed not only large scale fatigue tests using carbon & low-alloy steel flat plates and austenitic stainless steel piping but also fatigue tests using small specimens to obtain not only basic data but also fatigue data of mean stress effect.

✓The fatigue lives of not only the small specimens but also theCS & LAS plates and the stainless steel pipes are close tothe best-fit curve developed by the DFC1 subcommittee, andthe size effect can be considered as negligible.

✓The mean stress effect is remarkable in materials with highertensile strength. The correction of mean stress effect with theS-W-T approach shows good agreement with the BFCs.

18© 2016 MITSUBISHI HEAVY INDUSTRIES, LTD. All Rights Reserved.

Documents

Development of New Design Fatigue Curves in Japan · 2018. 9. 25. · Development of New Design Fatigue Curves in Japan ... data but also fatigue data of mean stress effect. •Fatigue