See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/313888618

FATIGUE CRACK GROWTH OF PIPELINE STEELS IN

GASEOUS HYDROGEN-PREDICTIVE MODEL
CALIBRATED TO API-5L X52*

Conference Paper · September 2012

CITATIONS READS

9 500

5 authors, including:

Robert Lewis Amaro Elizabeth S Drexler

National Institute of Standards and Technology, Boulde… National Institute of Standards and Technology
30 PUBLICATIONS 642 CITATIONS 64 PUBLICATIONS 1,156 CITATIONS

SEE PROFILE SEE PROFILE

Yaakov Levy Andrew J Slifka

Ben-Gurion University of the Negev National Institute of Standards and Technology
16 PUBLICATIONS 353 CITATIONS 85 PUBLICATIONS 1,411 CITATIONS

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Symposium on Hydrogen Embrittlement Understanding and Future Research Framework (ECF22 conference, Serbia, 26-
31 August 2018). View project

Hydrogen-assisted fatigue View project

All content following this page was uploaded by Andrew J Slifka on 22 February 2017.

The user has requested enhancement of the downloaded file.

 1

FATIGUE CRACK GROWTH OF PIPELINE STEELS IN GASEOUS

HYDROGEN- PREDICTIVE MODEL CALIBRATED TO API-5L X52*

ROBERT L. AMARO ELIZABETH S. DREXLER

National Institute of Standards and National Institute of Standards and
Technology, USA Technology, USA
NEHA RUSTAGI NICHOLAS E. NANNINGA
National Institute of Standards and National Institute of Standards and
Technology, USA Technology, USA
YAAKOV S. LEVY ANDREW J. SLIFKA
National Institute of Standards and National Institute of Standards and
Technology, USA Technology, USA

*Contribution of NIST, an agency of the US government. This material is declared a

work of the U.S. Government and is not subject to copyright protection in the United
States. Approved for public release; distribution is unlimited.

ABSTRACT
A phenomenological fatigue crack growth (FCG) model was previously
proposed to predict the FCG of API-5L X100 steel in high-pressure gaseous
hydrogen [1]. The model presumes that total FCG is a summation of a “pure”
fatigue component and a hydrogen-assisted (HA) component. The hydrogen-
assisted fatigue component is further divided into two regimes. The first
regime is dominated by the stress-assisted hydrogen concentration within the
fatigue process zone, while the second regime is controlled by the far-field
steady-state hydrogen concentration. This work outlines the process entailed in
model calibration to X52 steel in gaseous hydrogen. Model applicability to
multiple materials, having differing microstructure and characteristic length
scales, will be discussed in light of the model parameters and potential FCG
mechanisms.

INTRODUCTION
Hydrogen is envisioned as a key component of a future sustainable energy
society. Whether used as a “greening” agent in natural gas (through hydrogen
doping) [2], or as an energy carrier, hydrogen transport via steel pipelines is the
most likely method of transportation [3]. However, hydrogen adversely affects
the fatigue and fracture properties of carbon steel pipelines [4]. As such,
extensive experimental data and predictive models are required to enable the
design and installation of steel pipelines prior to the realization of a nationwide
hydrogen transportation system.
A fatigue crack growth (FCG) model was created to predict crack
propagation in X100 steels exposed to high-pressure hydrogen as a function of
hydrogen pressure and applied driving force [1]. The Stage II FCG model,
( ) EQ. 1
 2

is based upon the hypothesis that the total FCG, , is a result of the
superposition of fatigue in air (or an inert environment), , and

hydrogen-assisted (HA) fatigue, . The Dirac delta function, δ, is used to

activate the hydrogen-assisted component of the model when the hydrogen
pressure, PH, is greater than the threshold value, =0.02 MPa [4]. A
superposition model of similar form was initially proposed [5, 6] for corrosion
fatigue and has been subsequently used to correlate environmental fatigue crack
growth in oxidizing environments [7, 8], corrosive environments [9], high-
temperature environments inducing viscoplastic deformation [7],
thermomechanical fatigue [10], and crack advance segregated by microstructural
considerations [8]. The inert, Stage 2, fatigue response is governed by the
traditional Paris relationship,
EQ. 1
where A and b are material specific constants and ΔK is the stress intensity
range. The hydrogen-assisted FCG within the Stage 2, or Paris regime,

[( ) ( ) ] , EQ. 2

is modeled as a competition between a hydrogen-dominated component,

and a component dominated primarily by the FCG driving force, ,

henceforth referred to as transient and steady-state HA FCG, respectively. The
mathematical structure of the HA FCG module is predicated upon the
understanding that the FCG is a consequence of the interaction between the two
competing mechanisms. Whereas the hydrogen-dominated FCG governs the
material response in the transient regime (lower da/dN values), the driving-force
(ΔK) dominated FCG governs in the steady-state regime (higher da/dN values).
The transient HA FCG is defined as
( )
( ) , EQ. 3
where a1, B1, m1 and d1 are constants, Q is an activation energy taken as 27.1
kJ/mol-K [11], V = 2.0x10-6 m3/mol [12] is the partial molar volume of hydrogen
in the metal, σh is the maximum hydrostatic stress in the crack path in front of
the crack tip, R is the universal gas constant and T is the absolute temperature.
The hydrostatic stress is determined by use of the Hutchinson, Rice and
Rosengren crack tip solution [13, 14] and depends upon the material-specific
hardening exponent, n, and constant α. The two constants are determined by
fitting the Ramberg-Osgood uniaxial deformation response equation to tensile
test data [15]. The steady-state HA FCG is defined as
( )
( ) , EQ. 4
 3

where a2, B2, m2 and d2 are constants, σh is the hydrostatic stress in the crack
path in front of the crack tip at a distance equal to the previous cycles’ (or time
steps’) crack advance (da/dN), and all other variables are defined above.
This work details the steps involved in calibrating the model to a different
material (X52) and discusses the potential mechanisms responsible for
calibration differences between materials.

MATERIALS/EXPERIMENTAL PROCEDURE
Fatigue crack growth experiments were performed on API-5L X52 and
X100 pipeline steels in high-pressure hydrogen at the hydrogen laboratory of the
National Institute of Standards and technology in Boulder, Colorado. The
nominal chemical composition of the materials tested is provided in Table 1.

Table 1: Nominal chemical composition of API steels tested

C Mn Si S P Ni Cr Mo Nb+V+Ti Fe
X100 0.064 1.870 0.099 <0.001 0.009 0.470 0.023 0.230 0.036 Bal
X52 0.060 0.087 0.120 0.006 0.011 0.020 0.030 - 0.036 Bal

Compact tension specimens in the transverse-longitudinal orientation were

tested on a closed-loop 22 kip (100 kN) servo-hydraulic load frame. High-
pressure hydrogen tests were conducted at 250 psi (1.72 MPa), 1000 psi (6.89
MPa), 3000 psi (20.68 MPa) and 7000 psi (48.26 MPa). Tests performed in
gaseous hydrogen were conducted in a 20 ksi (138 MPa) pressure vessel. Test
control was handled by a commercially available software package with built-in
load control functionality. All tests were performed per ASTM E647-08 [16].
Feedback for the force loop was provided by a 10 kip (44 kN) proving ring
located inside the pressure vessel. The proving ring had a 5 lbf (0.02 kN)
resolution. Crack mouth opening displacement was monitored via clip gage
having a 3 mm range and 0.001 mm resolution. Hydrogen gas was either
hydrolyzed from ultrapure water or provided in “commercially pure” form from
an outside vendor. Hydrogen purity was tested periodically by an outside
vendor. Hydrogen pressure was maintained during testing via a program written
in-house that uses commercially available software. All test results reported
here were conducted at a load ratio of Rf = 0.5 and a frequency of 1 Hz. The
X100 specimens tested were all machined from the same section of pipe, i.e., the
same heat, as was the case for the X52 specimens.

TEST RESULTS
Fatigue crack test data for X52 and X100 tested in air and high-pressure
hydrogen are provided in Fig 1.
 4

0.1 0.01
Air Air
0.01 1.72 MPa 6.89 MPa
da/dN (mm/cycle)

da/dN (mm/cycle)
6.89 MPa 0.001
20.68 MPa
0.001 20.68 MPa
48.26 MPa
0.0001
0.0001

0.00001
0.00001

0.000001 0.000001
1 10 100 1 10 100
ΔK (Mpa-m1/2) ΔK (Mpa-m1/2)
Figure 1: Fatigue crack growth results for API-5L X100 steel (left) and API-
5L X52 steel (right) as a function of environment, defined in key as air or
the hydrogen test pressure.

The results of the X100 FCG tests indicate that as the hydrogen pressure is
decreased, and at lower ΔK values, the HA response approaches the FCG
response in air. This phenomenon was initially characterized as stress-corrosion
fatigue [17]. Furthermore, the results indicate that an increase in the hydrogen
test pressure yields an increase in the FCG rate. This observation can be seen by
the resulting increase in crack growth per cycle, at a given ΔK, as the hydrogen
pressure increases. Finally, the test results in hydrogen appear to converge at
higher crack-growth rates, regardless of hydrogen pressure. Similar results are
presented in [18, 19]. This trend indicates a minimization of the effect that the
hydrogen pressure has on the steady-state HA FCG regime.
The X52 data exhibit similar trends to that of the X100, with the exception
of the specimen tested in 7000 psi (48.62 MPa) hydrogen. This data set crosses
the other HA data near the “knee point”, or transition, between the transient
response and steady-state response. The 7000 psi hydrogen data appear to be an
anomaly, as the majority of the literature indicates a convergence of the FCG
response in this regime and not a cross-over.

MODEL CALIBRATION
The HA FCG model presented above was initially developed and calibrated
to predict the response of X100 steel. All model parameters were utilized to
predict the material response in the original formulation. Table 2 provides the
model parameter values calibrated to X100.

Table 2: Full model calibration for API-5L X100 steel

API-5L X100
A 9.9 x 10-9 b 2.83
Transient Steady State
a1 1.5 x 10-4 a2 8 x 10-4
B1 7.96 B2 3.15
m1 0.25 m2 0.1
d1 3 d2 1
 5

In this work a simplified version of the model will be used in order to elucidate
the material-specific model parameters that require calibration. The simplified
form sets d2 = 0 and B2 = b. As a result, the steady-state HA material response
is assumed to be unaffected by changes in hydrogen pressure. This assumption
is supported by the experimental results indicating that the FCG response of a
material tested in hydrogen converges to a single line in the steady-state regime
(higher values of da/dN), regardless of hydrogen pressure. A physical
interpretation of this model formulation is that the mechanism governing the
steady-state regime is dominated primarily by fatigue. In such a case case, when
gaseous hydrogen is present, the FCG response results from a competition
between a hydrogen-enhanced mechanism (transient regime) and a fatigue-
dominated mechanism (steady-state regime). At low da/dN, the crack extension
per cycle occurs within the region of material having the highest hydrogen
concentration, resulting from a peak in hydrostatic stress at some distance in
front of the crack tip. As such the FCG response is dominated by hydrogen-
assisted fatigue mechanisms. At higher values of da/dN, the crack extension per
cycle extends beyond the location of stress-enhanced hydrogen concentration.
In which case the first portion of the steady-state crack extension (per cycle) is
dominated by the increased hydrogen concentration in the fatigue process zone,
while the subsequent portion of the crack extension (per cycle) occurs as a result
of the underlying fatigue mechanism overpowering the effect of the far-field
hydrogen concentration within the material.
The simplified model was calibrated for each material as follows: (1) the
Paris relationship (EQ. 1) was fit to data created in air to determine constants A
and b. (2) The Paris relationship was then fit to the linear section of the transient
HA FCG response for all hydrogen pressures tested. The parameter B1 was set
equal to the average of the exponents within this regime. (3) The exponent, m1,
was then adjusted to account for the effect that the hydrogen pressure has upon
the transient FCG response. (4) The parameter B2 was then set equal to b. (5)
Lastly, the parameters a1 and a2 were adjusted for final fit. Calibration
constants for the simplified version of the model calibrated to X100 and X52 are
provided in Table 3.

Table 3: Simplified model calibration for API-5L X100 and X52 steel
API-5L X52 API-5L X100
A 7.1 x 10-9 b 2.9 A 9.9 x 10-9 b 2.83
Transient Steady State Transient Steady State
a1 1.2 x 10-15 a2 3.4 x 10-7 a1 3 x 10-13 a2 4 x 10-7
B1 10.24 B2 2.9 B1 7.96 B2 2.83
m1 0.7 m2 N/A m1 0.7 m2 N/A
d1 1 d2 0 d1 1 d2 0

With the exception of B1, the parameter values are very similar for both
materials. Experimental and correlated FCG results (determined by use of the
model parameters in Table 3) are provided in Figs. 2 and 3.
 6

0.01 Air
6.89 MPa
Correlation 6.89 MPa
0.001 20.68 MPa

da/dN (mm/cycle)
Correlation 20.68 MPa
48.26 MPa
0.0001 Correlation 48.26 MPa

0.00001

0.000001
1 10 100
ΔK (Mpa-m1/2)
Figure 2: Fatigue crack growth results and model correlations for API-5L
X52 steel as a function of environment, defined in key as air or the
hydrogen test pressure.

0.1
Air
Correlation Air
0.01 1.72 MPa
da/dN (mm/cycle)

Correlation 1.72 MPa

0.001 Correlation 6.89 MPa

20.68 MPa
Correlation 20.68 MPa
0.0001
6.89 MPa

0.00001

0.000001
1 10 100
ΔK (Mpa-m1/2)
Figure 3: Fatigue crack growth results and model correlations for API-5L
X100 steel as a function of environment

The model, even in simplified form, performs exceptionally well at

correlating the FCG response of both materials in air and high-pressure
hydrogen.

DISCUSSION

Note that the primary difference in parameter calibration between the

two materials’ occurs in the transient regime, specifically in variables a1 and B1.
As both materials’ HA FCG response false within the same da/dN-ΔK space the
value of a1 is viewed as subservient to that of B1. Given that the variable B1
governs the slope of the transient HA FCG regime, this variable is likely to shed
 7

light upon the material-specific hydrogen-FCG interactions. Particularly, it

accounts for the effect that the stress intensity has upon the FCG response.
However, the stress intensity is simply a proxy for the state of stress at the crack
tip. One will note that the FCG response of both materials in air is nearly
identical, indicating that the difference in material microstructure and
composition has little or no effect upon the inert FCG response. The differences
in the HA FCG response of the two materials may indicate a microstructural or
compositional effect upon hydrogen accumulation-deformation interactions in
the materials. The X52 microstructure tested has a characteristic length scale on
the order of 10 μm, and was composed primarily of ferrite and pearlite, while
the X100 microstructure has a characteristic length scale on the order of 1 μm
and was composed primarily of bainite and acicular ferrite [20]. Given that
grain boundaries serve as hydrogen-trapping sites, as well as dislocation pinning
locations, grain size would greatly affect the amount of accumulated hydrogen
within the material, and therefore the HA deformation response. The effect of
grain size and microstructure upon FCG response was examined in a study of
two X52 alloys [21]. The work looked at the FCG response of a 1960’s vintage
and 2010’s vintage alloy and found a substantial difference in HA FCG response
between the two alloys. Though the alloys are both technically X52, the older
alloy is comprised of ferrite-pearlite microstructure having a characteristic
length scale on the order of 5 μm. The newer alloy microstructure was
comprised of a mixture of polygonal and acicular ferrite, and had a characteristic
length scale on the order of 1 μm. Though the study did not pinpoint the effects
of grain size or microstructure, it provides insight to the effect of varying these
material properties. In a separate study, a bainitic microstructure has been
shown to increase a material’s susceptibility to hydrogen-enhanced damage [22,
23]. On the other hand, a study of two X60 candidate materials having different
microstructures concluded that the steel having a bainite/martensite
microstructure exhibited improved hydrogen deformation response when
compared to the material composed primarily of a ferrite-pearlite microstructure
[24]. A detailed set of parametric studies focusing on a single candidate
material having differing grain sizes and identical microstructure, as well as
identical grain size and differing microstructures, is required in order to
determine how these parameters influence HA FCG response.

CONCLUSIONS
This paper reviews an existing HA FCG model for pipeline steels and
outlines the steps required to calibrate the model to different grade steels. The
primary conclusions from this work are as follows: (1) The HA FCG model in a
simplified form performs well at correlating test results. (2) The simplified form
of the model assumes that a change in hydrogen pressure has minimal effect at
higher FCG rates, thereby implying that this material response regime has
saturated with respect to HA deformation mechanisms. (3) Model calibration to
different pipeline steels is straightforward. (4) A single model parameter stands
out as being material-hydrogen interaction specific. More research is required to
determine the basis for this parameter’s material-hydrogen specific nature.
 8

ACKNOWLEDGEMENTS
A portion of this research was funded by the U.S. Department of
Transportation. This research was performed while the first author held a
National Research Council Research Associateship Award at the National
Institute of Standards and Technology.

REFERENCES

[1] Amaro, R.L., Findley, K. O., Rustagi, N., Drexler, E. S., Slifka, A. J., 2013. "Fatigue Crack
growth of X100 Pipeline Steels in Gaseous Hydrogen," International Journal of Fatigue, in
Review.
[2] Naturalhy, 2010, "Preparing for the Hydrogen Economy by Using Existing Natural Gas System
as a Catalyst- Final Publishable Activity Report." Naturalhy. p. 68,
http://www.naturalhy.net/docs/project_reports/Final_Publishable_Activity_Report.pdf.
[3] Öney, F., Vezirolu, T. N., Dülger, Z., 1994. "Evaluation of pipeline transportation of hydrogen
and natural gas mixtures," International Journal of Hydrogen Energy, 19(10): p. 813-822.
[4] Somerday, B., P., 2008, "Technical Reference on Hydrogen Compatibility of Materials- Plain
Carbon Ferritic Steels: C-Mn Alloys (Code 1100)." Sandia National Laboratories, Livermore,
CA. p. 32.
[5] Bucci, R.J., 1970, "Environment enhanced fatigue and stress corrosion cracking of a titanium
alloy plus a simple model for the assessment of environmental influence on fatigue behavior."
Lehigh University, PhD: Bethlehem.
[6] Wei, R., P., Landes, J., 1970. "Correlation Between Sustained Load and fatigue Crack Growth in
High Strength Steels," Materials Research and Standards, 9: p. 25-27, 44-46.
[7] Miller, M.P., McDowell, D. L., Oehmke, R. L. T., 1992. "A Creep-Fatigue-Oxidation
Microcrack Propagation Model for Thermomechanical Fatigue," Journal of Engineering
Materials and Technology, 114(3): p. 282-288.
[8] Martínez-Esnaola, J.M., Martín-Meizoso, A., Affeldt, E. E., Bennett, A., Fuentes, M., 1997.
"Hight Temperature Fatigue in Single Crystal Superalloys," Fatigue & Fracture of Engineering
Materials & Structures, 20(5): p. 771-788.
[9] Wei, R.P., Gao, Ming, 1983. "Reconsideration of the superposition model for environmentally
assisted fatigue crack growth," Scripta Metallurgica, 17(7): p. 959-962.
[10] Maier, H., Teteruk, R., Christ, H., 2000. "Modeling thermomechanical fatigue life of high-
temperature titanium alloy IMI 834," Metallurgical and Materials Transactions A, 31(2): p.
431-444.
[11] Nelson, H., G., Stein, J., E., 1973, "Gas-Phase Hydrogen Permeation Through Alpha Iron, 4130
Steel, and 304 Stainless Steel From Less Than 100C to Near 600C." NASA Technical Note D-
7265.
[12] Hirth, J., 1980. "Effects of hydrogen on the properties of iron and steel," Metallurgical and
Materials Transactions A, 11(6): p. 861-890.
[13] Hutchinson, J.W., 1968. "Singular behaviour at the end of a tensile crack in a hardening
material," Journal of the Mechanics and Physics of Solids, 16(1): p. 13-31.
[14] Rice, J.R., Rosengren, G. F., 1968. "Plane strain deformation near a crack tip in a power-law
hardening material," Journal of the Mechanics and Physics of Solids, 16(1): p. 1-12.
[15] Ramberg, W., Osgood, W. R., 1943, "Description of Stress-Strain Curve by Three Parameters,"
National Advisory Committee for Aeronautics, Washington, DC. p. Note 902.
[16] 2008, "ASTM Standard E 647-08 "Test Method for Measurement of Fatigue Crack Growth
Rates"." ASTM International. p. 45.
[17] McEvily, A.J., Wei, R., P., 1971, in Corrosion Fatigue: Chemistry, Mechanics and
Microstructure, O. Devereux, McEvily, A.J., Staehle, R.W., Editor. NACE: Houston, Texas. p.
381-395.
[18] Stalheim, D., San Marchi, C., Somerday, B., Boggess, T., Jansto, S., Muralidharan, G.,
Sofronis, P. 2010. "Microstructure and Mechanical Property performance of Commercial
Grade API Pipeline Steels in High Pressure Gaseous Hydrogen'. in 8th International Pipeline
Conference. Calgary, Alberta, Canada: ASME.
 9

[19] San Marchi, C., Stalheim, D., G., Somerday, B., P., Boggess, T., Nibur, K., A., Jansto, S., 2010,
"Fracture and Fatigue of Commercial Grade API Pipeline Steels in Gaseous Hydrogen," in
ASME Pressure Vessels & Piping Division/ K-PVP 2010 Conference. ASME: Bellevue,
Washington, USA. p. 10.
[20] Nanninga, N.E., Levy, Y. S., Drexler, E. S., Condon, R. T., Stevenson, A. E., Slifka, A. J.,
2012. "Comparison of hydrogen embrittlement in three pipeline steels in high pressure gaseous
hydrogen environments," Corrosion Science, 59(0): p. 1-9.
[21] Slifka, A.J., Drexler, E. S., Rustagi, N., Lauria, D. S., McColskey, D. S., Amaro, R. L.,
Stalheim, D. G., Hayden, L. 2012. "Fatigue Crack Growth of Two X52 Pipeline Steels in a
Pressurized Hydrogen Environment'. in International Hydrogen Conference (to be presented).
Moran, WY, USA.
[22] Huang, F., Liu, J., Deng, Z. J., Cheng, J. H., Lu, Z. H., Li, X. G., 2010. "Effect of
microstructure and inclusions on hydrogen induced cracking susceptibility and hydrogen
trapping efficiency of X120 pipeline steel," Materials Science and Engineering: A, 527(26): p.
6997-7001.
[23] Arafin, M.A., Szpunar, J. A., 2011. "Effect of bainitic microstructure on the susceptibility of
pipeline steels to hydrogen induced cracking," Materials Science and Engineering: A, 528(15):
p. 4927-4940.
[24] Carneiro, R.A., Ratnapuli, R. C., de Freitas Cunha Lins, V., 2003. "The influence of chemical
composition and microstructure of API linepipe steels on hydrogen induced cracking and
sulfide stress corrosion cracking," Materials Science and Engineering: A, 357(1–2): p. 104-
110.

View publication stats