Fatigue Crack Growth and Crack Closure

Analyses on Thin ESE(T) and Thin/Thick C(T)

Specimens made of A36 Steel *

J. C. Newman, Jr.
Fatigue & Fracture Associates, LLC
Eupora, MS, USA

Dan Lingenfelser
HBM nCode Federal, LLC
Starkville, MS, USA

WebEx Meeting
Recent Advances on FCG Investigations and Modeling
22 May 2021

ffa * Work supported by HBM Prenscia HBM Prenscia

 Major Crack-Shielding Mechanisms

Plastic wake Oxide debris

(a) Plasticity-induced (b) Roughness-induced (c) Oxide/corrosion product-

closure closure induced closure
Modeling:
Crack contact and Crack contact Crack contact
residual stresses
are linked
May 2021 # 2
 Outline

• Problem statement
• A36 Steel from SAE FDE total-life project
• Low strength & very ductile
• Difference in crack-growth rates (~ 10x) at 12 MPa-m1/2
• Specimen and loading are within ASTM E-647 limits
• 19-mm thick C(T) specimen versus 6.35-mm thick ESE(T) specimen
• TCLR, CPCA, CPLR testing

• Test program to confirm and understand the differences

• Test results on thin and thick A36 specimens
• Crack-closure analyses (measurements and modeling)
• Conclusions

May 2021 # 3
Previous Fatigue-Crack-Growth Test Data Generated on A36 Steel

Specimen: FDE MSU

Type: Compact, C(T) ESE(T)
Thickness: 19 mm 6.35 mm

Width: 76 mm 38 mm
Test Procedure: Tensile-compression Compression-compression
pre-cracking, pre-cracking, constant-
constant amplitude amplitude (CPCA) or load
(TCCA) reduction (CPLR)

May 2021 # 4
Problem: Fatigue-Crack Growth Test Data on A36 Steel Differed

10-6
A36 Steel
R = 0.1

TCCA
TCLR
10-7
MSU: ESE(T) FDE: C(T)
B = 6.35 mm B = 19 mm
w = 38 mm w = 76 mm
dc/dN 10-8
m/cycle
CPCA

CPLR ?
10-9

10-10
4 5 6 7 8 9 10 15 20 30 40 50
1/2
DK, MPa-m
May 2021 # 5
Possible Reasons for Differences in Test Data on A36 Steel

• Specimen type: C(T) versus ESE(T)

• Specimen size: width and thickness

• Fatigue pre-cracking and test procedures

• Oxide debris accumulation on crack surfaces

• ???

May 2021 # 6
 ESE(T) Specimen and Pin-hole Modification

(a) Standard drilled pin holes (b) Beveled pin holes

May 2021 # 7
Compression Pre-cracking Constant-Amplitude and Constant-K
Loading

CPCA or CPCK Loading Pcp

|Kcp|/E ~ 0.005 √mm

Kcp = -28 MPa√m

Pcp = -9 kN (-2 kips)

Pmax ~ 1.3 to 4.5 kN

Pcp
May 2021 # 8
Compression Pre-cracking Load-Reduction and Constant-
Amplitude Loading

CPLR and CA Loading

May 2021 # 9
Residual Stress-Intensity Factors in A36 Steel

May 2021 # 10
New Test Program on A36 Steel Compact Specimens

Specimen Beveled
Width, mm
Thickness, Load Ratio, Test
Number Pin Holes mm R Method:
S1 No 76 19 0.1 CPCA
S2 No (a) 76 19 0.1 / 0.5 CPCA / CA
S3 Yes (b) 76 19 0.5 CPCA
S4 Yes (b) 76 19 0.1 CPCA
S5 Yes (b) 76 6.35 0.1 CPCA
S6 Yes (b) 76 6.35 0.5 CPCA

(a) Non-straight crack front due to out-of-plane bending

(b) Straight crack front due to uniform pin-hole loading

May 2021 # 11
Compact Specimen with Beveled Pin Holes and Backface Strain Gage

May 2021 # 12
 Comparison of Local and Remote Gage Measurements

Yoshiki Yamada (2005)

Same amount of noise
P/Pmax Local gage Remote gage
1.0 Better signal on local gage
Near BFS 7075-T651
crack tip C(T) Po_Local > Po_Remote
0.9
B=11.2 mm
0.8 W=76 mm

0.7 R=0.1
DK=13 MPa-m1/2
0.6 Approximation for Elber’s
crack-opening-load ratio
0.5 from remote method:
0.4
FASTRAN FTA readings OP0 = 2 OP1 – OP2
0.354 0.322, OP1
0.3
0.284, OP2
0.245, OP4 OP0 = 0.360
0.2
0.205, OP8
0.174, OP16 (only good for low-R tests)
0.1

0.0
Reduced strain May 2021 # 13
Fatigue-Crack Growth Test Data at R = 0.1 on A36 Steel C(T) Specimens

10-6
A36 Steel
R = 0.1

TCCA
TCLR
10-7
MSU: ESE(T) CP
B = 6.35 mm Criteria FDE: C(T)
w = 38 mm B = 19 mm
w = 76 mm
Invalid CP data
dc/dN 10-8
m/cycle
CPCA #S1
C(T)
CPLR B = 19 mm
10-9 w = 76 mm

10-10
4 5 6 7 8 9 10 15 20 30 40 50
1/2
DK, MPa-m
May 2021 # 14
Measured Crack-Opening-Load Ratio on “Thick” C(T) Specimen at R = 0.1

1.0
A36 Steel
C(T) (#S1)
B = 19 mm
R = 0.1
0.8

0.6
Po / Pmax
ci / w FASTRAN
0.4
a = 2.1
a = 2.5
a = 3.0
0.2 Major effect for cracks
a b c under bending (plane-
No Light Very light No apparent strain behavior) in
debris debris debris debris Region 3
0.0
0.2 0.3 0.4 0.5 0.6 0.7 0.8
c/w
May 2021 # 15
Fatigue-Crack-Growth Test on Non-Beveled C(T) Specimen

10-6
A36 Steel
R = 0.1

TCCA
TCLR
10-7
MSU: ESE(T)
B = 6.35 mm FDE: C(T)
w = 38 mm B = 19 mm
w = 76 mm
dc/dN 10-8
m/cycle
CPCA
#S2
C(T)
CPLR B = 19 mm
10-9 w = 76 mm
CPCA
Very dark debris on crack
faces and non-straight crack
front (non-bevel pin holes)
10-10
4 5 6 7 8 9 10 15 20 30 40 50
1/2
DK, MPa-m
May 2021 # 16
Effects of Beveling Pin Holes on C(T) Specimen

#S2 Non-beveled pin holes (R = 0.1 / 0.5)

#S4 Beveled pin holes (R = 0.1)

May 2021 # 17
Crack Face Debris caused Strange Fatigue-Crack-Growth Rate Behavior

10-6
A36 Steel
R = 0.1

TCCA
TCLR
10-7 MSU: ESE(T)
B = 6.35 mm FDE: C(T)
w = 38 mm B = 19 mm
CP w = 76 mm
Criteria
dc/dN 10-8 No apparent
debris
m/cycle
CPCA
#S4
C(T)
B = 19 mm
CPLR Very light w = 76 mm
debris
10-9

Very dark debris on crack

faces and straight crack
front (bevel pin holes)

10-10
4 5 6 7 8 9 10 15 20 30 40 50
1/2
DK, MPa-m
May 2021 # 18
Measured Crack-Opening-Load Ratio on “Thick” C(T) Specimen at R = 0.1 (2)

1.0
A36 Steel
C(T) (#S4)
B = 19 mm
R = 0.1
0.8

0.6

Po / Pmax
ci / w FASTRAN
0.4
a = 2.1
a = 2.5

0.2
a b c
No Very dark Light No apparent
debris debris debris debris
0.0
0.2 0.3 0.4 0.5 0.6 0.7
c/w
May 2021 # 19
Fatigue-Crack Growth Test Data at R = 0.1 on A36 Steel C(T) Specimen (2)

10-6
A36 Steel
R = 0.1

TCCA
TCLR
10-7 MSU: ESE(T)
B = 6.35 mm CP
Criteria FDE: C(T)
w = 38 mm
B = 19 mm
CP w = 76 mm
Criteria
dc/dN 10-8
m/cycle
CPCA
C(T)
CPLR B = 19 mm
10-9 w = 76 mm

10-10
4 5 6 7 8 9 10 15 20 30 40 50
1/2
DK, MPa-m
May 2021 # 20
Fatigue-Crack-Growth Test Data on “Thin” A36 Steel C(T) Specimen

10-6
A36 Steel
B = 6.35 mm
R = 0.1

10-7

MSU: ESE(T)
CP
Criteria
dc/dN 10-8
mm/cycle Invalid CP MSU: C(T)
Data
b
Light
10-9 a debris

10-10
5 6 7 8 9 10 15 20 30 40 50
1/2
DK, MPa-m
May 2021 # 21
Measured Crack-Opening-Load Ratio on “Thin” C(T) Specimen at R = 0.1

#S5 R = 01
1.0
A36 Steel
C(T) (#S5)
B = 6.35 mm
R = 0.1
0.8

#S6 R = 0.5
0.6

Po / Pmax
ci / w FASTRAN
0.4
a = 2.1
a = 2.5
a = 3.0
0.2
a b
No Light No apparent
debris debris debris
0.0
0.2 0.3 0.4 0.5 0.6 0.7 0.8
c/w
May 2021 # 22
Problem: Fatigue-Crack Growth Test Data on Two Steels Differed

A36 Steel
R = 0.5
10-7

TCCA
TCLR

FDE: C(T)
10-8 B = 19 mm
w = 76 mm
dc/dN
m/cycle
McKeighan et al
TC-128 Steel
10-9 C(T)
B = 6.35 mm
R = 0.5
or greater

10-10
1.5 2 3 4 5 6 8 10 15 20 30 40

DK, MPa-m1/2
May 2021 # 23
Problem: Fatigue-Crack Growth Test Data on A36 Steel at High R Differed

A36 Steel
R = 0.5
10-7

TCCA
TCLR

MSU: ESE(T)
B = 6.35 mm FDE: C(T)
10-8 w = 38 mm B = 19 mm
w = 76 mm
dc/dN CPCA
m/cycle
McKeighan et al
TC-128 Steel
10-9 C(T)
B = 6.35 mm
R = 0.5
or greater CPLR

10-10
1.5 2 3 4 5 6 8 10 15 20 30 40
1/2
DK, MPa-m
May 2021 # 24
Fatigue-Crack-Growth Test Data on Thick and Thin Specimens Agreed

A36 Steel
R = 0.5
10-7

TCLR
TCCA
MSU: ESE(T) Test data using
B = 6.35 mm FDE: C(T)
10-8 B = 19 mm
TCCA test method
w = 38 mm
w = 76 mm had most likely
dc/dN CPCA debris effects at
m/cycle No apparent low rates that did
McKeighan et al
debris
TC-128 Steel
CPCA not occur using
10-9 C(T)
B = 6.35 mm C(T) CPCA loading (?)
R = 0.5 B = 19 mm
or greater CPLR w = 76 mm

10-10
1.5 2 3 4 5 6 8 10 15 20 30 40
1/2
DK, MPa-m
May 2021 # 25
Effective Stress-Intensity Factor against Rate Relation on Thin and Thick
A36 and TC-128 Steel Specimens

A36 Steel
R = 0.1, 0.5, 0.7
10-7
DKeff (B = 6.35 mm)
Measured (OP0)
S5; R = 0.1

10-8 DKeff (B = 19 mm)

Measured (OP0)
S4; R = 0.1
dc/dN
m/cycle
McKeighan et al
TC-128 Steel
10-9 C(T) MSU: C(T)
B = 6.35 mm
B = 19 mm; R = 0.5
R = 0.5
or greater MSU: ESE(T)
B = 6.35 mm; R = 0.7
10-10
1.5 2 3 4 5 6 8 10 15 20 30 40

DKeff, MPa-m1/2
May 2021 # 26
 Effective Stress-Intensity Factor against Rate Relation on Thin and Thick
A36 and TC-128 Steel Specimens (2)

A36 Steel
R = 0.1, 0.5, 0.7
10-7
DKeff (B = 6.35 mm)
Measured (OP0)
S5; R = 0.1

10-8 DKeff (B = 19 mm)

Measured (OP0)
S4; R = 0.1
dc/dN
m/cycle
McKeighan et al
TC-128 Steel
10-9 C(T) MSU: C(T)
B = 6.35 mm
B = 19 mm; R = 0.5
R = 0.5
or greater MSU: ESE(T)
B = 6.35 mm; R = 0.7
10-10
1.5 2 3 4 5 6 8 10 15 20 30 40

DKeff, MPa-m1/2
May 2021 # 27
Crack-Face Appearance and Methods to Determine Crack-Closure Effects

#S1 R = 0.1 (debris and plasticity)

Measured
crack-opening loads
(OP0 = 2 OP1 – OP2)
E-647 – Appendix X2

Calculated
crack-opening loads
(FASTRAN a = 2.5)

#S3 R = 0.5 (plasticity)

May 2021 # 28
FASTRAN – Calculated (Plasticity) Crack-Opening-Load Ratios

FASTRAN: a = 2.5 (Plane-strain conditions)

R Po/Pmax U
0.7 0.7 1.0
0.6 0.61 0.99
0.5 0.52 0.96
0.1 0.31 0.77

Elber: DKeff = U DK

May 2021 # 29
Fatigue-Crack-Growth Test Data on Thick FDE Specimen Disagreed

A36 Steel
R = 0.1, 0.5, 0.7
10-7
DKeff (B = 6.35 mm)
Measured (OP0)
S5; R = 0.1 TCLR
TCCA
Crack-opening
10-8 DKeff (B = 19 mm) loads measured
Measured (OP0) during TCCA test
S4; R = 0.1
dc/dN FDE: C(T) with crack-
B = 19 mm
m/cycle
R = 0.5
surface debris
McKeighan et al
TC-128 Steel effects may have
10-9 C(T) MSU: C(T) explained the
B = 6.35 mm
B = 19 mm; R = 0.5 disagreement!
R = 0.5
or greater MSU: ESE(T)
B = 6.35 mm; R = 0.7
10-10
1.5 2 3 4 5 6 8 10 15 20 30 40

DKeff, MPa-m1/2
May 2021 # 30
Crack-Opening Loads Measured on CPCK Test on Thin ESE(T)
Specimen at R = 0.1

May 2021 # 31
Fatigue-Crack-Surface during CPCK Test at R = 0.1

EDM Compression
Notch Pre-cracking Debris Debris

May 2021 # 32
Attempt to Conduct a Constant DK Test on A36 Steel

May 2021 # 33
Deep-Crack Bend Specimens under Plane-Strain Conditions

Solanki et al (2003)

TC-128B
(R = 0.6)

A36 &
TC-128B

May 2021 # 34
Tension Specimens under Plane-Strain Conditions Exhibit Closure

Solanki et al (2004)

Load history and debris!

May 2021 # 35
 Conclusions on A36 Steel Fatigue-Crack-Growth Tests
1. Non-beveled pin holes in ESE(T) and C(T) specimens caused non-straight
crack fronts, while beveled pin holes produced straight crack fronts.
2. For low load ratios (R = 0.1), the thick steel C(T) specimens caused more
crack face “debris” than thin specimens (more crack closure).
3. For low load ratios (R = 0.1), the DK-rate data for thin and thick C(T)
specimens differed greatly due to debris.
4. For high load ratios (R = 0.5), the thin and thick C(T) specimens showed
fairly clean crack faces (less debris and less crack closure).
5. For high load ratios (R = 0.5), the DK-rate data for thin and thick C(T)
specimens agreed, but the TCCA generated test data differed due to
unknown causes.
6. Deep cracks in bending, ESE(T) and C(T), are under more plane-strain
conditions and less crack closure (no R-shift at high rates), but deep
cracks in tension, M(T), may have crack closure and an R-shift.
7. A crack-closure analysis using measured crack-opening loads for R = 0.1
and calculated values for high load ratios (R = 0.5 or greater) produced a
nearly unique DKeff-rate relation.

ffa HBM Prenscia

May 2021 # 36